diff options
author | Dimitry Andric <dim@FreeBSD.org> | 2015-05-27 18:44:32 +0000 |
---|---|---|
committer | Dimitry Andric <dim@FreeBSD.org> | 2015-05-27 18:44:32 +0000 |
commit | 5a5ac124e1efaf208671f01c46edb15f29ed2a0b (patch) | |
tree | a6140557876943cdd800ee997c9317283394b22c /lib/Bitcode | |
parent | f03b5bed27d0d2eafd68562ce14f8b5e3f1f0801 (diff) | |
download | src-5a5ac124e1efaf208671f01c46edb15f29ed2a0b.tar.gz src-5a5ac124e1efaf208671f01c46edb15f29ed2a0b.zip |
Notes
Diffstat (limited to 'lib/Bitcode')
-rw-r--r-- | lib/Bitcode/Reader/BitcodeReader.cpp | 1267 | ||||
-rw-r--r-- | lib/Bitcode/Reader/BitcodeReader.h | 365 | ||||
-rw-r--r-- | lib/Bitcode/Reader/BitstreamReader.cpp | 40 | ||||
-rw-r--r-- | lib/Bitcode/Reader/CMakeLists.txt | 3 | ||||
-rw-r--r-- | lib/Bitcode/Writer/BitcodeWriter.cpp | 595 | ||||
-rw-r--r-- | lib/Bitcode/Writer/BitcodeWriterPass.cpp | 16 | ||||
-rw-r--r-- | lib/Bitcode/Writer/ValueEnumerator.cpp | 39 | ||||
-rw-r--r-- | lib/Bitcode/Writer/ValueEnumerator.h | 25 |
8 files changed, 1698 insertions, 652 deletions
diff --git a/lib/Bitcode/Reader/BitcodeReader.cpp b/lib/Bitcode/Reader/BitcodeReader.cpp index 3d9546fc2b79..f6d5ccc1a59e 100644 --- a/lib/Bitcode/Reader/BitcodeReader.cpp +++ b/lib/Bitcode/Reader/BitcodeReader.cpp @@ -8,32 +8,380 @@ //===----------------------------------------------------------------------===// #include "llvm/Bitcode/ReaderWriter.h" -#include "BitcodeReader.h" +#include "llvm/ADT/STLExtras.h" #include "llvm/ADT/SmallString.h" #include "llvm/ADT/SmallVector.h" +#include "llvm/ADT/Triple.h" +#include "llvm/Bitcode/BitstreamReader.h" #include "llvm/Bitcode/LLVMBitCodes.h" #include "llvm/IR/AutoUpgrade.h" #include "llvm/IR/Constants.h" +#include "llvm/IR/DebugInfo.h" +#include "llvm/IR/DebugInfoMetadata.h" #include "llvm/IR/DerivedTypes.h" #include "llvm/IR/DiagnosticPrinter.h" +#include "llvm/IR/GVMaterializer.h" #include "llvm/IR/InlineAsm.h" #include "llvm/IR/IntrinsicInst.h" #include "llvm/IR/LLVMContext.h" #include "llvm/IR/Module.h" #include "llvm/IR/OperandTraits.h" #include "llvm/IR/Operator.h" +#include "llvm/IR/ValueHandle.h" #include "llvm/Support/DataStream.h" #include "llvm/Support/ManagedStatic.h" #include "llvm/Support/MathExtras.h" #include "llvm/Support/MemoryBuffer.h" #include "llvm/Support/raw_ostream.h" - +#include <deque> using namespace llvm; +namespace { enum { SWITCH_INST_MAGIC = 0x4B5 // May 2012 => 1205 => Hex }; +class BitcodeReaderValueList { + std::vector<WeakVH> ValuePtrs; + + /// ResolveConstants - As we resolve forward-referenced constants, we add + /// information about them to this vector. This allows us to resolve them in + /// bulk instead of resolving each reference at a time. See the code in + /// ResolveConstantForwardRefs for more information about this. + /// + /// The key of this vector is the placeholder constant, the value is the slot + /// number that holds the resolved value. + typedef std::vector<std::pair<Constant*, unsigned> > ResolveConstantsTy; + ResolveConstantsTy ResolveConstants; + LLVMContext &Context; +public: + BitcodeReaderValueList(LLVMContext &C) : Context(C) {} + ~BitcodeReaderValueList() { + assert(ResolveConstants.empty() && "Constants not resolved?"); + } + + // vector compatibility methods + unsigned size() const { return ValuePtrs.size(); } + void resize(unsigned N) { ValuePtrs.resize(N); } + void push_back(Value *V) { + ValuePtrs.push_back(V); + } + + void clear() { + assert(ResolveConstants.empty() && "Constants not resolved?"); + ValuePtrs.clear(); + } + + Value *operator[](unsigned i) const { + assert(i < ValuePtrs.size()); + return ValuePtrs[i]; + } + + Value *back() const { return ValuePtrs.back(); } + void pop_back() { ValuePtrs.pop_back(); } + bool empty() const { return ValuePtrs.empty(); } + void shrinkTo(unsigned N) { + assert(N <= size() && "Invalid shrinkTo request!"); + ValuePtrs.resize(N); + } + + Constant *getConstantFwdRef(unsigned Idx, Type *Ty); + Value *getValueFwdRef(unsigned Idx, Type *Ty); + + void AssignValue(Value *V, unsigned Idx); + + /// ResolveConstantForwardRefs - Once all constants are read, this method bulk + /// resolves any forward references. + void ResolveConstantForwardRefs(); +}; + +class BitcodeReaderMDValueList { + unsigned NumFwdRefs; + bool AnyFwdRefs; + unsigned MinFwdRef; + unsigned MaxFwdRef; + std::vector<TrackingMDRef> MDValuePtrs; + + LLVMContext &Context; +public: + BitcodeReaderMDValueList(LLVMContext &C) + : NumFwdRefs(0), AnyFwdRefs(false), Context(C) {} + + // vector compatibility methods + unsigned size() const { return MDValuePtrs.size(); } + void resize(unsigned N) { MDValuePtrs.resize(N); } + void push_back(Metadata *MD) { MDValuePtrs.emplace_back(MD); } + void clear() { MDValuePtrs.clear(); } + Metadata *back() const { return MDValuePtrs.back(); } + void pop_back() { MDValuePtrs.pop_back(); } + bool empty() const { return MDValuePtrs.empty(); } + + Metadata *operator[](unsigned i) const { + assert(i < MDValuePtrs.size()); + return MDValuePtrs[i]; + } + + void shrinkTo(unsigned N) { + assert(N <= size() && "Invalid shrinkTo request!"); + MDValuePtrs.resize(N); + } + + Metadata *getValueFwdRef(unsigned Idx); + void AssignValue(Metadata *MD, unsigned Idx); + void tryToResolveCycles(); +}; + +class BitcodeReader : public GVMaterializer { + LLVMContext &Context; + DiagnosticHandlerFunction DiagnosticHandler; + Module *TheModule; + std::unique_ptr<MemoryBuffer> Buffer; + std::unique_ptr<BitstreamReader> StreamFile; + BitstreamCursor Stream; + DataStreamer *LazyStreamer; + uint64_t NextUnreadBit; + bool SeenValueSymbolTable; + + std::vector<Type*> TypeList; + BitcodeReaderValueList ValueList; + BitcodeReaderMDValueList MDValueList; + std::vector<Comdat *> ComdatList; + SmallVector<Instruction *, 64> InstructionList; + + std::vector<std::pair<GlobalVariable*, unsigned> > GlobalInits; + std::vector<std::pair<GlobalAlias*, unsigned> > AliasInits; + std::vector<std::pair<Function*, unsigned> > FunctionPrefixes; + std::vector<std::pair<Function*, unsigned> > FunctionPrologues; + + SmallVector<Instruction*, 64> InstsWithTBAATag; + + /// MAttributes - The set of attributes by index. Index zero in the + /// file is for null, and is thus not represented here. As such all indices + /// are off by one. + std::vector<AttributeSet> MAttributes; + + /// \brief The set of attribute groups. + std::map<unsigned, AttributeSet> MAttributeGroups; + + /// FunctionBBs - While parsing a function body, this is a list of the basic + /// blocks for the function. + std::vector<BasicBlock*> FunctionBBs; + + // When reading the module header, this list is populated with functions that + // have bodies later in the file. + std::vector<Function*> FunctionsWithBodies; + + // When intrinsic functions are encountered which require upgrading they are + // stored here with their replacement function. + typedef std::vector<std::pair<Function*, Function*> > UpgradedIntrinsicMap; + UpgradedIntrinsicMap UpgradedIntrinsics; + + // Map the bitcode's custom MDKind ID to the Module's MDKind ID. + DenseMap<unsigned, unsigned> MDKindMap; + + // Several operations happen after the module header has been read, but + // before function bodies are processed. This keeps track of whether + // we've done this yet. + bool SeenFirstFunctionBody; + + /// DeferredFunctionInfo - When function bodies are initially scanned, this + /// map contains info about where to find deferred function body in the + /// stream. + DenseMap<Function*, uint64_t> DeferredFunctionInfo; + + /// When Metadata block is initially scanned when parsing the module, we may + /// choose to defer parsing of the metadata. This vector contains info about + /// which Metadata blocks are deferred. + std::vector<uint64_t> DeferredMetadataInfo; + + /// These are basic blocks forward-referenced by block addresses. They are + /// inserted lazily into functions when they're loaded. The basic block ID is + /// its index into the vector. + DenseMap<Function *, std::vector<BasicBlock *>> BasicBlockFwdRefs; + std::deque<Function *> BasicBlockFwdRefQueue; + + /// UseRelativeIDs - Indicates that we are using a new encoding for + /// instruction operands where most operands in the current + /// FUNCTION_BLOCK are encoded relative to the instruction number, + /// for a more compact encoding. Some instruction operands are not + /// relative to the instruction ID: basic block numbers, and types. + /// Once the old style function blocks have been phased out, we would + /// not need this flag. + bool UseRelativeIDs; + + /// True if all functions will be materialized, negating the need to process + /// (e.g.) blockaddress forward references. + bool WillMaterializeAllForwardRefs; + + /// Functions that have block addresses taken. This is usually empty. + SmallPtrSet<const Function *, 4> BlockAddressesTaken; + + /// True if any Metadata block has been materialized. + bool IsMetadataMaterialized; + + bool StripDebugInfo = false; + +public: + std::error_code Error(BitcodeError E, const Twine &Message); + std::error_code Error(BitcodeError E); + std::error_code Error(const Twine &Message); + + explicit BitcodeReader(MemoryBuffer *buffer, LLVMContext &C, + DiagnosticHandlerFunction DiagnosticHandler); + explicit BitcodeReader(DataStreamer *streamer, LLVMContext &C, + DiagnosticHandlerFunction DiagnosticHandler); + ~BitcodeReader() override { FreeState(); } + + std::error_code materializeForwardReferencedFunctions(); + + void FreeState(); + + void releaseBuffer(); + + bool isDematerializable(const GlobalValue *GV) const override; + std::error_code materialize(GlobalValue *GV) override; + std::error_code materializeModule(Module *M) override; + std::vector<StructType *> getIdentifiedStructTypes() const override; + void dematerialize(GlobalValue *GV) override; + + /// @brief Main interface to parsing a bitcode buffer. + /// @returns true if an error occurred. + std::error_code ParseBitcodeInto(Module *M, + bool ShouldLazyLoadMetadata = false); + + /// @brief Cheap mechanism to just extract module triple + /// @returns true if an error occurred. + ErrorOr<std::string> parseTriple(); + + static uint64_t decodeSignRotatedValue(uint64_t V); + + /// Materialize any deferred Metadata block. + std::error_code materializeMetadata() override; + + void setStripDebugInfo() override; + +private: + std::vector<StructType *> IdentifiedStructTypes; + StructType *createIdentifiedStructType(LLVMContext &Context, StringRef Name); + StructType *createIdentifiedStructType(LLVMContext &Context); + + Type *getTypeByID(unsigned ID); + Value *getFnValueByID(unsigned ID, Type *Ty) { + if (Ty && Ty->isMetadataTy()) + return MetadataAsValue::get(Ty->getContext(), getFnMetadataByID(ID)); + return ValueList.getValueFwdRef(ID, Ty); + } + Metadata *getFnMetadataByID(unsigned ID) { + return MDValueList.getValueFwdRef(ID); + } + BasicBlock *getBasicBlock(unsigned ID) const { + if (ID >= FunctionBBs.size()) return nullptr; // Invalid ID + return FunctionBBs[ID]; + } + AttributeSet getAttributes(unsigned i) const { + if (i-1 < MAttributes.size()) + return MAttributes[i-1]; + return AttributeSet(); + } + + /// getValueTypePair - Read a value/type pair out of the specified record from + /// slot 'Slot'. Increment Slot past the number of slots used in the record. + /// Return true on failure. + bool getValueTypePair(SmallVectorImpl<uint64_t> &Record, unsigned &Slot, + unsigned InstNum, Value *&ResVal) { + if (Slot == Record.size()) return true; + unsigned ValNo = (unsigned)Record[Slot++]; + // Adjust the ValNo, if it was encoded relative to the InstNum. + if (UseRelativeIDs) + ValNo = InstNum - ValNo; + if (ValNo < InstNum) { + // If this is not a forward reference, just return the value we already + // have. + ResVal = getFnValueByID(ValNo, nullptr); + return ResVal == nullptr; + } + if (Slot == Record.size()) + return true; + + unsigned TypeNo = (unsigned)Record[Slot++]; + ResVal = getFnValueByID(ValNo, getTypeByID(TypeNo)); + return ResVal == nullptr; + } + + /// popValue - Read a value out of the specified record from slot 'Slot'. + /// Increment Slot past the number of slots used by the value in the record. + /// Return true if there is an error. + bool popValue(SmallVectorImpl<uint64_t> &Record, unsigned &Slot, + unsigned InstNum, Type *Ty, Value *&ResVal) { + if (getValue(Record, Slot, InstNum, Ty, ResVal)) + return true; + // All values currently take a single record slot. + ++Slot; + return false; + } + + /// getValue -- Like popValue, but does not increment the Slot number. + bool getValue(SmallVectorImpl<uint64_t> &Record, unsigned Slot, + unsigned InstNum, Type *Ty, Value *&ResVal) { + ResVal = getValue(Record, Slot, InstNum, Ty); + return ResVal == nullptr; + } + + /// getValue -- Version of getValue that returns ResVal directly, + /// or 0 if there is an error. + Value *getValue(SmallVectorImpl<uint64_t> &Record, unsigned Slot, + unsigned InstNum, Type *Ty) { + if (Slot == Record.size()) return nullptr; + unsigned ValNo = (unsigned)Record[Slot]; + // Adjust the ValNo, if it was encoded relative to the InstNum. + if (UseRelativeIDs) + ValNo = InstNum - ValNo; + return getFnValueByID(ValNo, Ty); + } + + /// getValueSigned -- Like getValue, but decodes signed VBRs. + Value *getValueSigned(SmallVectorImpl<uint64_t> &Record, unsigned Slot, + unsigned InstNum, Type *Ty) { + if (Slot == Record.size()) return nullptr; + unsigned ValNo = (unsigned)decodeSignRotatedValue(Record[Slot]); + // Adjust the ValNo, if it was encoded relative to the InstNum. + if (UseRelativeIDs) + ValNo = InstNum - ValNo; + return getFnValueByID(ValNo, Ty); + } + + /// Converts alignment exponent (i.e. power of two (or zero)) to the + /// corresponding alignment to use. If alignment is too large, returns + /// a corresponding error code. + std::error_code parseAlignmentValue(uint64_t Exponent, unsigned &Alignment); + std::error_code ParseAttrKind(uint64_t Code, Attribute::AttrKind *Kind); + std::error_code ParseModule(bool Resume, bool ShouldLazyLoadMetadata = false); + std::error_code ParseAttributeBlock(); + std::error_code ParseAttributeGroupBlock(); + std::error_code ParseTypeTable(); + std::error_code ParseTypeTableBody(); + + std::error_code ParseValueSymbolTable(); + std::error_code ParseConstants(); + std::error_code RememberAndSkipFunctionBody(); + /// Save the positions of the Metadata blocks and skip parsing the blocks. + std::error_code rememberAndSkipMetadata(); + std::error_code ParseFunctionBody(Function *F); + std::error_code GlobalCleanup(); + std::error_code ResolveGlobalAndAliasInits(); + std::error_code ParseMetadata(); + std::error_code ParseMetadataAttachment(Function &F); + ErrorOr<std::string> parseModuleTriple(); + std::error_code ParseUseLists(); + std::error_code InitStream(); + std::error_code InitStreamFromBuffer(); + std::error_code InitLazyStream(); + std::error_code FindFunctionInStream( + Function *F, + DenseMap<Function *, uint64_t>::iterator DeferredFunctionInfoIterator); +}; +} // namespace + BitcodeDiagnosticInfo::BitcodeDiagnosticInfo(std::error_code EC, DiagnosticSeverity Severity, const Twine &Msg) @@ -53,6 +401,12 @@ static std::error_code Error(DiagnosticHandlerFunction DiagnosticHandler, return Error(DiagnosticHandler, EC, EC.message()); } +static std::error_code Error(DiagnosticHandlerFunction DiagnosticHandler, + const Twine &Message) { + return Error(DiagnosticHandler, + make_error_code(BitcodeError::CorruptedBitcode), Message); +} + std::error_code BitcodeReader::Error(BitcodeError E, const Twine &Message) { return ::Error(DiagnosticHandler, make_error_code(E), Message); } @@ -79,7 +433,7 @@ BitcodeReader::BitcodeReader(MemoryBuffer *buffer, LLVMContext &C, TheModule(nullptr), Buffer(buffer), LazyStreamer(nullptr), NextUnreadBit(0), SeenValueSymbolTable(false), ValueList(C), MDValueList(C), SeenFirstFunctionBody(false), UseRelativeIDs(false), - WillMaterializeAllForwardRefs(false) {} + WillMaterializeAllForwardRefs(false), IsMetadataMaterialized(false) {} BitcodeReader::BitcodeReader(DataStreamer *streamer, LLVMContext &C, DiagnosticHandlerFunction DiagnosticHandler) @@ -87,7 +441,7 @@ BitcodeReader::BitcodeReader(DataStreamer *streamer, LLVMContext &C, TheModule(nullptr), Buffer(nullptr), LazyStreamer(streamer), NextUnreadBit(0), SeenValueSymbolTable(false), ValueList(C), MDValueList(C), SeenFirstFunctionBody(false), UseRelativeIDs(false), - WillMaterializeAllForwardRefs(false) {} + WillMaterializeAllForwardRefs(false), IsMetadataMaterialized(false) {} std::error_code BitcodeReader::materializeForwardReferencedFunctions() { if (WillMaterializeAllForwardRefs) @@ -133,6 +487,7 @@ void BitcodeReader::FreeState() { std::vector<BasicBlock*>().swap(FunctionBBs); std::vector<Function*>().swap(FunctionsWithBodies); DeferredFunctionInfo.clear(); + DeferredMetadataInfo.clear(); MDKindMap.clear(); assert(BasicBlockFwdRefs.empty() && "Unresolved blockaddress fwd references"); @@ -156,19 +511,27 @@ static bool ConvertToString(ArrayRef<uint64_t> Record, unsigned Idx, return false; } +static bool hasImplicitComdat(size_t Val) { + switch (Val) { + default: + return false; + case 1: // Old WeakAnyLinkage + case 4: // Old LinkOnceAnyLinkage + case 10: // Old WeakODRLinkage + case 11: // Old LinkOnceODRLinkage + return true; + } +} + static GlobalValue::LinkageTypes getDecodedLinkage(unsigned Val) { switch (Val) { default: // Map unknown/new linkages to external case 0: return GlobalValue::ExternalLinkage; - case 1: - return GlobalValue::WeakAnyLinkage; case 2: return GlobalValue::AppendingLinkage; case 3: return GlobalValue::InternalLinkage; - case 4: - return GlobalValue::LinkOnceAnyLinkage; case 5: return GlobalValue::ExternalLinkage; // Obsolete DLLImportLinkage case 6: @@ -179,10 +542,6 @@ static GlobalValue::LinkageTypes getDecodedLinkage(unsigned Val) { return GlobalValue::CommonLinkage; case 9: return GlobalValue::PrivateLinkage; - case 10: - return GlobalValue::WeakODRLinkage; - case 11: - return GlobalValue::LinkOnceODRLinkage; case 12: return GlobalValue::AvailableExternallyLinkage; case 13: @@ -191,6 +550,18 @@ static GlobalValue::LinkageTypes getDecodedLinkage(unsigned Val) { return GlobalValue::PrivateLinkage; // Obsolete LinkerPrivateWeakLinkage case 15: return GlobalValue::ExternalLinkage; // Obsolete LinkOnceODRAutoHideLinkage + case 1: // Old value with implicit comdat. + case 16: + return GlobalValue::WeakAnyLinkage; + case 10: // Old value with implicit comdat. + case 17: + return GlobalValue::WeakODRLinkage; + case 4: // Old value with implicit comdat. + case 18: + return GlobalValue::LinkOnceAnyLinkage; + case 11: // Old value with implicit comdat. + case 19: + return GlobalValue::LinkOnceODRLinkage; } } @@ -242,27 +613,42 @@ static int GetDecodedCastOpcode(unsigned Val) { case bitc::CAST_ADDRSPACECAST: return Instruction::AddrSpaceCast; } } + static int GetDecodedBinaryOpcode(unsigned Val, Type *Ty) { + bool IsFP = Ty->isFPOrFPVectorTy(); + // BinOps are only valid for int/fp or vector of int/fp types + if (!IsFP && !Ty->isIntOrIntVectorTy()) + return -1; + switch (Val) { - default: return -1; + default: + return -1; case bitc::BINOP_ADD: - return Ty->isFPOrFPVectorTy() ? Instruction::FAdd : Instruction::Add; + return IsFP ? Instruction::FAdd : Instruction::Add; case bitc::BINOP_SUB: - return Ty->isFPOrFPVectorTy() ? Instruction::FSub : Instruction::Sub; + return IsFP ? Instruction::FSub : Instruction::Sub; case bitc::BINOP_MUL: - return Ty->isFPOrFPVectorTy() ? Instruction::FMul : Instruction::Mul; - case bitc::BINOP_UDIV: return Instruction::UDiv; + return IsFP ? Instruction::FMul : Instruction::Mul; + case bitc::BINOP_UDIV: + return IsFP ? -1 : Instruction::UDiv; case bitc::BINOP_SDIV: - return Ty->isFPOrFPVectorTy() ? Instruction::FDiv : Instruction::SDiv; - case bitc::BINOP_UREM: return Instruction::URem; + return IsFP ? Instruction::FDiv : Instruction::SDiv; + case bitc::BINOP_UREM: + return IsFP ? -1 : Instruction::URem; case bitc::BINOP_SREM: - return Ty->isFPOrFPVectorTy() ? Instruction::FRem : Instruction::SRem; - case bitc::BINOP_SHL: return Instruction::Shl; - case bitc::BINOP_LSHR: return Instruction::LShr; - case bitc::BINOP_ASHR: return Instruction::AShr; - case bitc::BINOP_AND: return Instruction::And; - case bitc::BINOP_OR: return Instruction::Or; - case bitc::BINOP_XOR: return Instruction::Xor; + return IsFP ? Instruction::FRem : Instruction::SRem; + case bitc::BINOP_SHL: + return IsFP ? -1 : Instruction::Shl; + case bitc::BINOP_LSHR: + return IsFP ? -1 : Instruction::LShr; + case bitc::BINOP_ASHR: + return IsFP ? -1 : Instruction::AShr; + case bitc::BINOP_AND: + return IsFP ? -1 : Instruction::And; + case bitc::BINOP_OR: + return IsFP ? -1 : Instruction::Or; + case bitc::BINOP_XOR: + return IsFP ? -1 : Instruction::Xor; } } @@ -332,7 +718,7 @@ namespace { /// @brief A class for maintaining the slot number definition /// as a placeholder for the actual definition for forward constants defs. class ConstantPlaceHolder : public ConstantExpr { - void operator=(const ConstantPlaceHolder &) LLVM_DELETED_FUNCTION; + void operator=(const ConstantPlaceHolder &) = delete; public: // allocate space for exactly one operand void *operator new(size_t s) { @@ -399,7 +785,8 @@ Constant *BitcodeReaderValueList::getConstantFwdRef(unsigned Idx, resize(Idx + 1); if (Value *V = ValuePtrs[Idx]) { - assert(Ty == V->getType() && "Type mismatch in constant table!"); + if (Ty != V->getType()) + report_fatal_error("Type mismatch in constant table!"); return cast<Constant>(V); } @@ -410,11 +797,17 @@ Constant *BitcodeReaderValueList::getConstantFwdRef(unsigned Idx, } Value *BitcodeReaderValueList::getValueFwdRef(unsigned Idx, Type *Ty) { + // Bail out for a clearly invalid value. This would make us call resize(0) + if (Idx == UINT_MAX) + return nullptr; + if (Idx >= size()) resize(Idx + 1); if (Value *V = ValuePtrs[Idx]) { - assert((!Ty || Ty == V->getType()) && "Type mismatch in value table!"); + // If the types don't match, it's invalid. + if (Ty && Ty != V->getType()) + return nullptr; return V; } @@ -525,9 +918,8 @@ void BitcodeReaderMDValueList::AssignValue(Metadata *MD, unsigned Idx) { } // If there was a forward reference to this value, replace it. - MDNodeFwdDecl *PrevMD = cast<MDNodeFwdDecl>(OldMD.get()); + TempMDTuple PrevMD(cast<MDTuple>(OldMD.get())); PrevMD->replaceAllUsesWith(MD); - MDNode::deleteTemporary(PrevMD); --NumFwdRefs; } @@ -549,7 +941,7 @@ Metadata *BitcodeReaderMDValueList::getValueFwdRef(unsigned Idx) { ++NumFwdRefs; // Create and return a placeholder, which will later be RAUW'd. - Metadata *MD = MDNode::getTemporary(Context, None); + Metadata *MD = MDNode::getTemporary(Context, None).release(); MDValuePtrs[Idx].reset(MD); return MD; } @@ -566,9 +958,12 @@ void BitcodeReaderMDValueList::tryToResolveCycles() { // Resolve any cycles. for (unsigned I = MinFwdRef, E = MaxFwdRef + 1; I != E; ++I) { auto &MD = MDValuePtrs[I]; - assert(!(MD && isa<MDNodeFwdDecl>(MD)) && "Unexpected forward reference"); - if (auto *N = dyn_cast_or_null<UniquableMDNode>(MD)) - N->resolveCycles(); + auto *N = dyn_cast_or_null<MDNode>(MD); + if (!N) + continue; + + assert(!N->isTemporary() && "Unexpected forward reference"); + N->resolveCycles(); } // Make sure we return early again until there's another forward ref. @@ -701,6 +1096,8 @@ static Attribute::AttrKind GetAttrFromCode(uint64_t Code) { return Attribute::InAlloca; case bitc::ATTR_KIND_COLD: return Attribute::Cold; + case bitc::ATTR_KIND_CONVERGENT: + return Attribute::Convergent; case bitc::ATTR_KIND_INLINE_HINT: return Attribute::InlineHint; case bitc::ATTR_KIND_IN_REG: @@ -731,6 +1128,8 @@ static Attribute::AttrKind GetAttrFromCode(uint64_t Code) { return Attribute::NonNull; case bitc::ATTR_KIND_DEREFERENCEABLE: return Attribute::Dereferenceable; + case bitc::ATTR_KIND_DEREFERENCEABLE_OR_NULL: + return Attribute::DereferenceableOrNull; case bitc::ATTR_KIND_NO_RED_ZONE: return Attribute::NoRedZone; case bitc::ATTR_KIND_NO_RETURN: @@ -774,6 +1173,16 @@ static Attribute::AttrKind GetAttrFromCode(uint64_t Code) { } } +std::error_code BitcodeReader::parseAlignmentValue(uint64_t Exponent, + unsigned &Alignment) { + // Note: Alignment in bitcode files is incremented by 1, so that zero + // can be used for default alignment. + if (Exponent > Value::MaxAlignmentExponent + 1) + return Error("Invalid alignment value"); + Alignment = (1 << static_cast<unsigned>(Exponent)) >> 1; + return std::error_code(); +} + std::error_code BitcodeReader::ParseAttrKind(uint64_t Code, Attribute::AttrKind *Kind) { *Kind = GetAttrFromCode(Code); @@ -837,6 +1246,8 @@ std::error_code BitcodeReader::ParseAttributeGroupBlock() { B.addStackAlignmentAttr(Record[++i]); else if (Kind == Attribute::Dereferenceable) B.addDereferenceableAttr(Record[++i]); + else if (Kind == Attribute::DereferenceableOrNull) + B.addDereferenceableOrNullAttr(Record[++i]); } else { // String attribute assert((Record[i] == 3 || Record[i] == 4) && "Invalid attribute group entry"); @@ -943,12 +1354,17 @@ std::error_code BitcodeReader::ParseTypeTableBody() { case bitc::TYPE_CODE_X86_MMX: // X86_MMX ResultTy = Type::getX86_MMXTy(Context); break; - case bitc::TYPE_CODE_INTEGER: // INTEGER: [width] + case bitc::TYPE_CODE_INTEGER: { // INTEGER: [width] if (Record.size() < 1) return Error("Invalid record"); - ResultTy = IntegerType::get(Context, Record[0]); + uint64_t NumBits = Record[0]; + if (NumBits < IntegerType::MIN_INT_BITS || + NumBits > IntegerType::MAX_INT_BITS) + return Error("Bitwidth for integer type out of range"); + ResultTy = IntegerType::get(Context, NumBits); break; + } case bitc::TYPE_CODE_POINTER: { // POINTER: [pointee type] or // [pointee type, address space] if (Record.size() < 1) @@ -957,7 +1373,8 @@ std::error_code BitcodeReader::ParseTypeTableBody() { if (Record.size() == 2) AddressSpace = Record[1]; ResultTy = getTypeByID(Record[0]); - if (!ResultTy) + if (!ResultTy || + !PointerType::isValidElementType(ResultTy)) return Error("Invalid type"); ResultTy = PointerType::get(ResultTy, AddressSpace); break; @@ -988,8 +1405,11 @@ std::error_code BitcodeReader::ParseTypeTableBody() { return Error("Invalid record"); SmallVector<Type*, 8> ArgTys; for (unsigned i = 2, e = Record.size(); i != e; ++i) { - if (Type *T = getTypeByID(Record[i])) + if (Type *T = getTypeByID(Record[i])) { + if (!FunctionType::isValidArgumentType(T)) + return Error("Invalid function argument type"); ArgTys.push_back(T); + } else break; } @@ -1071,25 +1491,27 @@ std::error_code BitcodeReader::ParseTypeTableBody() { case bitc::TYPE_CODE_ARRAY: // ARRAY: [numelts, eltty] if (Record.size() < 2) return Error("Invalid record"); - if ((ResultTy = getTypeByID(Record[1]))) - ResultTy = ArrayType::get(ResultTy, Record[0]); - else + ResultTy = getTypeByID(Record[1]); + if (!ResultTy || !ArrayType::isValidElementType(ResultTy)) return Error("Invalid type"); + ResultTy = ArrayType::get(ResultTy, Record[0]); break; case bitc::TYPE_CODE_VECTOR: // VECTOR: [numelts, eltty] if (Record.size() < 2) return Error("Invalid record"); - if ((ResultTy = getTypeByID(Record[1]))) - ResultTy = VectorType::get(ResultTy, Record[0]); - else + ResultTy = getTypeByID(Record[1]); + if (!ResultTy || !StructType::isValidElementType(ResultTy)) return Error("Invalid type"); + ResultTy = VectorType::get(ResultTy, Record[0]); break; } if (NumRecords >= TypeList.size()) return Error("Invalid TYPE table"); + if (TypeList[NumRecords]) + return Error( + "Invalid TYPE table: Only named structs can be forward referenced"); assert(ResultTy && "Didn't read a type?"); - assert(!TypeList[NumRecords] && "Already read type?"); TypeList[NumRecords++] = ResultTy; } } @@ -1100,6 +1522,8 @@ std::error_code BitcodeReader::ParseValueSymbolTable() { SmallVector<uint64_t, 64> Record; + Triple TT(TheModule->getTargetTriple()); + // Read all the records for this value table. SmallString<128> ValueName; while (1) { @@ -1130,6 +1554,14 @@ std::error_code BitcodeReader::ParseValueSymbolTable() { Value *V = ValueList[ValueID]; V->setName(StringRef(ValueName.data(), ValueName.size())); + if (auto *GO = dyn_cast<GlobalObject>(V)) { + if (GO->getComdat() == reinterpret_cast<Comdat *>(1)) { + if (TT.isOSBinFormatMachO()) + GO->setComdat(nullptr); + else + GO->setComdat(TheModule->getOrInsertComdat(V->getName())); + } + } ValueName.clear(); break; } @@ -1148,7 +1580,10 @@ std::error_code BitcodeReader::ParseValueSymbolTable() { } } +static int64_t unrotateSign(uint64_t U) { return U & 1 ? ~(U >> 1) : U >> 1; } + std::error_code BitcodeReader::ParseMetadata() { + IsMetadataMaterialized = true; unsigned NextMDValueNo = MDValueList.size(); if (Stream.EnterSubBlock(bitc::METADATA_BLOCK_ID)) @@ -1156,6 +1591,22 @@ std::error_code BitcodeReader::ParseMetadata() { SmallVector<uint64_t, 64> Record; + auto getMD = + [&](unsigned ID) -> Metadata *{ return MDValueList.getValueFwdRef(ID); }; + auto getMDOrNull = [&](unsigned ID) -> Metadata *{ + if (ID) + return getMD(ID - 1); + return nullptr; + }; + auto getMDString = [&](unsigned ID) -> MDString *{ + // This requires that the ID is not really a forward reference. In + // particular, the MDString must already have been resolved. + return cast_or_null<MDString>(getMDOrNull(ID)); + }; + +#define GET_OR_DISTINCT(CLASS, DISTINCT, ARGS) \ + (DISTINCT ? CLASS::getDistinct ARGS : CLASS::get ARGS) + // Read all the records. while (1) { BitstreamEntry Entry = Stream.advanceSkippingSubblocks(); @@ -1283,16 +1734,270 @@ std::error_code BitcodeReader::ParseMetadata() { if (Record.size() != 5) return Error("Invalid record"); - auto get = Record[0] ? MDLocation::getDistinct : MDLocation::get; unsigned Line = Record[1]; unsigned Column = Record[2]; MDNode *Scope = cast<MDNode>(MDValueList.getValueFwdRef(Record[3])); Metadata *InlinedAt = Record[4] ? MDValueList.getValueFwdRef(Record[4] - 1) : nullptr; - MDValueList.AssignValue(get(Context, Line, Column, Scope, InlinedAt), + MDValueList.AssignValue( + GET_OR_DISTINCT(DILocation, Record[0], + (Context, Line, Column, Scope, InlinedAt)), + NextMDValueNo++); + break; + } + case bitc::METADATA_GENERIC_DEBUG: { + if (Record.size() < 4) + return Error("Invalid record"); + + unsigned Tag = Record[1]; + unsigned Version = Record[2]; + + if (Tag >= 1u << 16 || Version != 0) + return Error("Invalid record"); + + auto *Header = getMDString(Record[3]); + SmallVector<Metadata *, 8> DwarfOps; + for (unsigned I = 4, E = Record.size(); I != E; ++I) + DwarfOps.push_back(Record[I] ? MDValueList.getValueFwdRef(Record[I] - 1) + : nullptr); + MDValueList.AssignValue(GET_OR_DISTINCT(GenericDINode, Record[0], + (Context, Tag, Header, DwarfOps)), + NextMDValueNo++); + break; + } + case bitc::METADATA_SUBRANGE: { + if (Record.size() != 3) + return Error("Invalid record"); + + MDValueList.AssignValue( + GET_OR_DISTINCT(DISubrange, Record[0], + (Context, Record[1], unrotateSign(Record[2]))), + NextMDValueNo++); + break; + } + case bitc::METADATA_ENUMERATOR: { + if (Record.size() != 3) + return Error("Invalid record"); + + MDValueList.AssignValue(GET_OR_DISTINCT(DIEnumerator, Record[0], + (Context, unrotateSign(Record[1]), + getMDString(Record[2]))), NextMDValueNo++); break; } + case bitc::METADATA_BASIC_TYPE: { + if (Record.size() != 6) + return Error("Invalid record"); + + MDValueList.AssignValue( + GET_OR_DISTINCT(DIBasicType, Record[0], + (Context, Record[1], getMDString(Record[2]), + Record[3], Record[4], Record[5])), + NextMDValueNo++); + break; + } + case bitc::METADATA_DERIVED_TYPE: { + if (Record.size() != 12) + return Error("Invalid record"); + + MDValueList.AssignValue( + GET_OR_DISTINCT(DIDerivedType, Record[0], + (Context, Record[1], getMDString(Record[2]), + getMDOrNull(Record[3]), Record[4], + getMDOrNull(Record[5]), getMDOrNull(Record[6]), + Record[7], Record[8], Record[9], Record[10], + getMDOrNull(Record[11]))), + NextMDValueNo++); + break; + } + case bitc::METADATA_COMPOSITE_TYPE: { + if (Record.size() != 16) + return Error("Invalid record"); + + MDValueList.AssignValue( + GET_OR_DISTINCT(DICompositeType, Record[0], + (Context, Record[1], getMDString(Record[2]), + getMDOrNull(Record[3]), Record[4], + getMDOrNull(Record[5]), getMDOrNull(Record[6]), + Record[7], Record[8], Record[9], Record[10], + getMDOrNull(Record[11]), Record[12], + getMDOrNull(Record[13]), getMDOrNull(Record[14]), + getMDString(Record[15]))), + NextMDValueNo++); + break; + } + case bitc::METADATA_SUBROUTINE_TYPE: { + if (Record.size() != 3) + return Error("Invalid record"); + + MDValueList.AssignValue( + GET_OR_DISTINCT(DISubroutineType, Record[0], + (Context, Record[1], getMDOrNull(Record[2]))), + NextMDValueNo++); + break; + } + case bitc::METADATA_FILE: { + if (Record.size() != 3) + return Error("Invalid record"); + + MDValueList.AssignValue( + GET_OR_DISTINCT(DIFile, Record[0], (Context, getMDString(Record[1]), + getMDString(Record[2]))), + NextMDValueNo++); + break; + } + case bitc::METADATA_COMPILE_UNIT: { + if (Record.size() < 14 || Record.size() > 15) + return Error("Invalid record"); + + MDValueList.AssignValue( + GET_OR_DISTINCT(DICompileUnit, Record[0], + (Context, Record[1], getMDOrNull(Record[2]), + getMDString(Record[3]), Record[4], + getMDString(Record[5]), Record[6], + getMDString(Record[7]), Record[8], + getMDOrNull(Record[9]), getMDOrNull(Record[10]), + getMDOrNull(Record[11]), getMDOrNull(Record[12]), + getMDOrNull(Record[13]), + Record.size() == 14 ? 0 : Record[14])), + NextMDValueNo++); + break; + } + case bitc::METADATA_SUBPROGRAM: { + if (Record.size() != 19) + return Error("Invalid record"); + + MDValueList.AssignValue( + GET_OR_DISTINCT( + DISubprogram, Record[0], + (Context, getMDOrNull(Record[1]), getMDString(Record[2]), + getMDString(Record[3]), getMDOrNull(Record[4]), Record[5], + getMDOrNull(Record[6]), Record[7], Record[8], Record[9], + getMDOrNull(Record[10]), Record[11], Record[12], Record[13], + Record[14], getMDOrNull(Record[15]), getMDOrNull(Record[16]), + getMDOrNull(Record[17]), getMDOrNull(Record[18]))), + NextMDValueNo++); + break; + } + case bitc::METADATA_LEXICAL_BLOCK: { + if (Record.size() != 5) + return Error("Invalid record"); + + MDValueList.AssignValue( + GET_OR_DISTINCT(DILexicalBlock, Record[0], + (Context, getMDOrNull(Record[1]), + getMDOrNull(Record[2]), Record[3], Record[4])), + NextMDValueNo++); + break; + } + case bitc::METADATA_LEXICAL_BLOCK_FILE: { + if (Record.size() != 4) + return Error("Invalid record"); + + MDValueList.AssignValue( + GET_OR_DISTINCT(DILexicalBlockFile, Record[0], + (Context, getMDOrNull(Record[1]), + getMDOrNull(Record[2]), Record[3])), + NextMDValueNo++); + break; + } + case bitc::METADATA_NAMESPACE: { + if (Record.size() != 5) + return Error("Invalid record"); + + MDValueList.AssignValue( + GET_OR_DISTINCT(DINamespace, Record[0], + (Context, getMDOrNull(Record[1]), + getMDOrNull(Record[2]), getMDString(Record[3]), + Record[4])), + NextMDValueNo++); + break; + } + case bitc::METADATA_TEMPLATE_TYPE: { + if (Record.size() != 3) + return Error("Invalid record"); + + MDValueList.AssignValue(GET_OR_DISTINCT(DITemplateTypeParameter, + Record[0], + (Context, getMDString(Record[1]), + getMDOrNull(Record[2]))), + NextMDValueNo++); + break; + } + case bitc::METADATA_TEMPLATE_VALUE: { + if (Record.size() != 5) + return Error("Invalid record"); + + MDValueList.AssignValue( + GET_OR_DISTINCT(DITemplateValueParameter, Record[0], + (Context, Record[1], getMDString(Record[2]), + getMDOrNull(Record[3]), getMDOrNull(Record[4]))), + NextMDValueNo++); + break; + } + case bitc::METADATA_GLOBAL_VAR: { + if (Record.size() != 11) + return Error("Invalid record"); + + MDValueList.AssignValue( + GET_OR_DISTINCT(DIGlobalVariable, Record[0], + (Context, getMDOrNull(Record[1]), + getMDString(Record[2]), getMDString(Record[3]), + getMDOrNull(Record[4]), Record[5], + getMDOrNull(Record[6]), Record[7], Record[8], + getMDOrNull(Record[9]), getMDOrNull(Record[10]))), + NextMDValueNo++); + break; + } + case bitc::METADATA_LOCAL_VAR: { + // 10th field is for the obseleted 'inlinedAt:' field. + if (Record.size() != 9 && Record.size() != 10) + return Error("Invalid record"); + + MDValueList.AssignValue( + GET_OR_DISTINCT(DILocalVariable, Record[0], + (Context, Record[1], getMDOrNull(Record[2]), + getMDString(Record[3]), getMDOrNull(Record[4]), + Record[5], getMDOrNull(Record[6]), Record[7], + Record[8])), + NextMDValueNo++); + break; + } + case bitc::METADATA_EXPRESSION: { + if (Record.size() < 1) + return Error("Invalid record"); + + MDValueList.AssignValue( + GET_OR_DISTINCT(DIExpression, Record[0], + (Context, makeArrayRef(Record).slice(1))), + NextMDValueNo++); + break; + } + case bitc::METADATA_OBJC_PROPERTY: { + if (Record.size() != 8) + return Error("Invalid record"); + + MDValueList.AssignValue( + GET_OR_DISTINCT(DIObjCProperty, Record[0], + (Context, getMDString(Record[1]), + getMDOrNull(Record[2]), Record[3], + getMDString(Record[4]), getMDString(Record[5]), + Record[6], getMDOrNull(Record[7]))), + NextMDValueNo++); + break; + } + case bitc::METADATA_IMPORTED_ENTITY: { + if (Record.size() != 6) + return Error("Invalid record"); + + MDValueList.AssignValue( + GET_OR_DISTINCT(DIImportedEntity, Record[0], + (Context, Record[1], getMDOrNull(Record[2]), + getMDOrNull(Record[3]), Record[4], + getMDString(Record[5]))), + NextMDValueNo++); + break; + } case bitc::METADATA_STRING: { std::string String(Record.begin(), Record.end()); llvm::UpgradeMDStringConstant(String); @@ -1314,6 +2019,7 @@ std::error_code BitcodeReader::ParseMetadata() { } } } +#undef GET_OR_DISTINCT } /// decodeSignRotatedValue - Decode a signed value stored with the sign bit in @@ -1635,19 +2341,29 @@ std::error_code BitcodeReader::ParseConstants() { } case bitc::CST_CODE_CE_INBOUNDS_GEP: case bitc::CST_CODE_CE_GEP: { // CE_GEP: [n x operands] - if (Record.size() & 1) - return Error("Invalid record"); + unsigned OpNum = 0; + Type *PointeeType = nullptr; + if (Record.size() % 2) + PointeeType = getTypeByID(Record[OpNum++]); SmallVector<Constant*, 16> Elts; - for (unsigned i = 0, e = Record.size(); i != e; i += 2) { - Type *ElTy = getTypeByID(Record[i]); + while (OpNum != Record.size()) { + Type *ElTy = getTypeByID(Record[OpNum++]); if (!ElTy) return Error("Invalid record"); - Elts.push_back(ValueList.getConstantFwdRef(Record[i+1], ElTy)); + Elts.push_back(ValueList.getConstantFwdRef(Record[OpNum++], ElTy)); } + + if (PointeeType && + PointeeType != + cast<SequentialType>(Elts[0]->getType()->getScalarType()) + ->getElementType()) + return Error("Explicit gep operator type does not match pointee type " + "of pointer operand"); + ArrayRef<Constant *> Indices(Elts.begin() + 1, Elts.end()); - V = ConstantExpr::getGetElementPtr(Elts[0], Indices, + V = ConstantExpr::getGetElementPtr(PointeeType, Elts[0], Indices, BitCode == - bitc::CST_CODE_CE_INBOUNDS_GEP); + bitc::CST_CODE_CE_INBOUNDS_GEP); break; } case bitc::CST_CODE_CE_SELECT: { // CE_SELECT: [opval#, opval#, opval#] @@ -1917,6 +2633,32 @@ std::error_code BitcodeReader::ParseUseLists() { } } +/// When we see the block for metadata, remember where it is and then skip it. +/// This lets us lazily deserialize the metadata. +std::error_code BitcodeReader::rememberAndSkipMetadata() { + // Save the current stream state. + uint64_t CurBit = Stream.GetCurrentBitNo(); + DeferredMetadataInfo.push_back(CurBit); + + // Skip over the block for now. + if (Stream.SkipBlock()) + return Error("Invalid record"); + return std::error_code(); +} + +std::error_code BitcodeReader::materializeMetadata() { + for (uint64_t BitPos : DeferredMetadataInfo) { + // Move the bit stream to the saved position. + Stream.JumpToBit(BitPos); + if (std::error_code EC = ParseMetadata()) + return EC; + } + DeferredMetadataInfo.clear(); + return std::error_code(); +} + +void BitcodeReader::setStripDebugInfo() { StripDebugInfo = true; } + /// RememberAndSkipFunctionBody - When we see the block for a function body, /// remember where it is and then skip it. This lets us lazily deserialize the /// functions. @@ -1967,7 +2709,8 @@ std::error_code BitcodeReader::GlobalCleanup() { return std::error_code(); } -std::error_code BitcodeReader::ParseModule(bool Resume) { +std::error_code BitcodeReader::ParseModule(bool Resume, + bool ShouldLazyLoadMetadata) { if (Resume) Stream.JumpToBit(NextUnreadBit); else if (Stream.EnterSubBlock(bitc::MODULE_BLOCK_ID)) @@ -2021,6 +2764,12 @@ std::error_code BitcodeReader::ParseModule(bool Resume) { return EC; break; case bitc::METADATA_BLOCK_ID: + if (ShouldLazyLoadMetadata && !IsMetadataMaterialized) { + if (std::error_code EC = rememberAndSkipMetadata()) + return EC; + break; + } + assert(DeferredMetadataInfo.empty() && "Unexpected deferred metadata"); if (std::error_code EC = ParseMetadata()) return EC; break; @@ -2139,21 +2888,31 @@ std::error_code BitcodeReader::ParseModule(bool Resume) { } // GLOBALVAR: [pointer type, isconst, initid, // linkage, alignment, section, visibility, threadlocal, - // unnamed_addr, dllstorageclass] + // unnamed_addr, externally_initialized, dllstorageclass, + // comdat] case bitc::MODULE_CODE_GLOBALVAR: { if (Record.size() < 6) return Error("Invalid record"); Type *Ty = getTypeByID(Record[0]); if (!Ty) return Error("Invalid record"); - if (!Ty->isPointerTy()) - return Error("Invalid type for value"); - unsigned AddressSpace = cast<PointerType>(Ty)->getAddressSpace(); - Ty = cast<PointerType>(Ty)->getElementType(); + bool isConstant = Record[1] & 1; + bool explicitType = Record[1] & 2; + unsigned AddressSpace; + if (explicitType) { + AddressSpace = Record[1] >> 2; + } else { + if (!Ty->isPointerTy()) + return Error("Invalid type for value"); + AddressSpace = cast<PointerType>(Ty)->getAddressSpace(); + Ty = cast<PointerType>(Ty)->getElementType(); + } - bool isConstant = Record[1]; - GlobalValue::LinkageTypes Linkage = getDecodedLinkage(Record[3]); - unsigned Alignment = (1 << Record[4]) >> 1; + uint64_t RawLinkage = Record[3]; + GlobalValue::LinkageTypes Linkage = getDecodedLinkage(RawLinkage); + unsigned Alignment; + if (std::error_code EC = parseAlignmentValue(Record[4], Alignment)) + return EC; std::string Section; if (Record[5]) { if (Record[5]-1 >= SectionTable.size()) @@ -2190,7 +2949,7 @@ std::error_code BitcodeReader::ParseModule(bool Resume) { if (Record.size() > 10) NewGV->setDLLStorageClass(GetDecodedDLLStorageClass(Record[10])); else - UpgradeDLLImportExportLinkage(NewGV, Record[3]); + UpgradeDLLImportExportLinkage(NewGV, RawLinkage); ValueList.push_back(NewGV); @@ -2198,11 +2957,15 @@ std::error_code BitcodeReader::ParseModule(bool Resume) { if (unsigned InitID = Record[2]) GlobalInits.push_back(std::make_pair(NewGV, InitID-1)); - if (Record.size() > 11) + if (Record.size() > 11) { if (unsigned ComdatID = Record[11]) { - assert(ComdatID <= ComdatList.size()); + if (ComdatID > ComdatList.size()) + return Error("Invalid global variable comdat ID"); NewGV->setComdat(ComdatList[ComdatID - 1]); } + } else if (hasImplicitComdat(RawLinkage)) { + NewGV->setComdat(reinterpret_cast<Comdat *>(1)); + } break; } // FUNCTION: [type, callingconv, isproto, linkage, paramattr, @@ -2214,10 +2977,9 @@ std::error_code BitcodeReader::ParseModule(bool Resume) { Type *Ty = getTypeByID(Record[0]); if (!Ty) return Error("Invalid record"); - if (!Ty->isPointerTy()) - return Error("Invalid type for value"); - FunctionType *FTy = - dyn_cast<FunctionType>(cast<PointerType>(Ty)->getElementType()); + if (auto *PTy = dyn_cast<PointerType>(Ty)) + Ty = PTy->getElementType(); + auto *FTy = dyn_cast<FunctionType>(Ty); if (!FTy) return Error("Invalid type for value"); @@ -2226,10 +2988,14 @@ std::error_code BitcodeReader::ParseModule(bool Resume) { Func->setCallingConv(static_cast<CallingConv::ID>(Record[1])); bool isProto = Record[2]; - Func->setLinkage(getDecodedLinkage(Record[3])); + uint64_t RawLinkage = Record[3]; + Func->setLinkage(getDecodedLinkage(RawLinkage)); Func->setAttributes(getAttributes(Record[4])); - Func->setAlignment((1 << Record[5]) >> 1); + unsigned Alignment; + if (std::error_code EC = parseAlignmentValue(Record[5], Alignment)) + return EC; + Func->setAlignment(Alignment); if (Record[6]) { if (Record[6]-1 >= SectionTable.size()) return Error("Invalid ID"); @@ -2240,7 +3006,7 @@ std::error_code BitcodeReader::ParseModule(bool Resume) { // FIXME: Change to an error if non-default in 4.0. Func->setVisibility(GetDecodedVisibility(Record[7])); if (Record.size() > 8 && Record[8]) { - if (Record[8]-1 > GCTable.size()) + if (Record[8]-1 >= GCTable.size()) return Error("Invalid ID"); Func->setGC(GCTable[Record[8]-1].c_str()); } @@ -2254,13 +3020,17 @@ std::error_code BitcodeReader::ParseModule(bool Resume) { if (Record.size() > 11) Func->setDLLStorageClass(GetDecodedDLLStorageClass(Record[11])); else - UpgradeDLLImportExportLinkage(Func, Record[3]); + UpgradeDLLImportExportLinkage(Func, RawLinkage); - if (Record.size() > 12) + if (Record.size() > 12) { if (unsigned ComdatID = Record[12]) { - assert(ComdatID <= ComdatList.size()); + if (ComdatID > ComdatList.size()) + return Error("Invalid function comdat ID"); Func->setComdat(ComdatList[ComdatID - 1]); } + } else if (hasImplicitComdat(RawLinkage)) { + Func->setComdat(reinterpret_cast<Comdat *>(1)); + } if (Record.size() > 13 && Record[13] != 0) FunctionPrefixes.push_back(std::make_pair(Func, Record[13]-1)); @@ -2290,8 +3060,7 @@ std::error_code BitcodeReader::ParseModule(bool Resume) { return Error("Invalid type for value"); auto *NewGA = - GlobalAlias::create(PTy->getElementType(), PTy->getAddressSpace(), - getDecodedLinkage(Record[2]), "", TheModule); + GlobalAlias::create(PTy, getDecodedLinkage(Record[2]), "", TheModule); // Old bitcode files didn't have visibility field. // Local linkage must have default visibility. if (Record.size() > 3 && !NewGA->hasLocalLinkage()) @@ -2321,7 +3090,8 @@ std::error_code BitcodeReader::ParseModule(bool Resume) { } } -std::error_code BitcodeReader::ParseBitcodeInto(Module *M) { +std::error_code BitcodeReader::ParseBitcodeInto(Module *M, + bool ShouldLazyLoadMetadata) { TheModule = nullptr; if (std::error_code EC = InitStream()) @@ -2339,8 +3109,12 @@ std::error_code BitcodeReader::ParseBitcodeInto(Module *M) { // We expect a number of well-defined blocks, though we don't necessarily // need to understand them all. while (1) { - if (Stream.AtEndOfStream()) - return std::error_code(); + if (Stream.AtEndOfStream()) { + if (TheModule) + return std::error_code(); + // We didn't really read a proper Module. + return Error("Malformed IR file"); + } BitstreamEntry Entry = Stream.advance(BitstreamCursor::AF_DontAutoprocessAbbrevs); @@ -2362,7 +3136,7 @@ std::error_code BitcodeReader::ParseBitcodeInto(Module *M) { if (TheModule) return Error("Invalid multiple blocks"); TheModule = M; - if (std::error_code EC = ParseModule(false)) + if (std::error_code EC = ParseModule(false, ShouldLazyLoadMetadata)) return EC; if (LazyStreamer) return std::error_code(); @@ -2468,7 +3242,7 @@ ErrorOr<std::string> BitcodeReader::parseTriple() { } /// ParseMetadataAttachment - Parse metadata attachments. -std::error_code BitcodeReader::ParseMetadataAttachment() { +std::error_code BitcodeReader::ParseMetadataAttachment(Function &F) { if (Stream.EnterSubBlock(bitc::METADATA_ATTACHMENT_ID)) return Error("Invalid record"); @@ -2494,8 +3268,21 @@ std::error_code BitcodeReader::ParseMetadataAttachment() { break; case bitc::METADATA_ATTACHMENT: { unsigned RecordLength = Record.size(); - if (Record.empty() || (RecordLength - 1) % 2 == 1) + if (Record.empty()) return Error("Invalid record"); + if (RecordLength % 2 == 0) { + // A function attachment. + for (unsigned I = 0; I != RecordLength; I += 2) { + auto K = MDKindMap.find(Record[I]); + if (K == MDKindMap.end()) + return Error("Invalid ID"); + Metadata *MD = MDValueList.getValueFwdRef(Record[I + 1]); + F.setMetadata(K->second, cast<MDNode>(MD)); + } + continue; + } + + // An instruction attachment. Instruction *Inst = InstructionList[Record[0]]; for (unsigned i = 1; i != RecordLength; i = i+2) { unsigned Kind = Record[i]; @@ -2518,6 +3305,20 @@ std::error_code BitcodeReader::ParseMetadataAttachment() { } } +static std::error_code TypeCheckLoadStoreInst(DiagnosticHandlerFunction DH, + Type *ValType, Type *PtrType) { + if (!isa<PointerType>(PtrType)) + return Error(DH, "Load/Store operand is not a pointer type"); + Type *ElemType = cast<PointerType>(PtrType)->getElementType(); + + if (ValType && ValType != ElemType) + return Error(DH, "Explicit load/store type does not match pointee type of " + "pointer operand"); + if (!PointerType::isLoadableOrStorableType(ElemType)) + return Error(DH, "Cannot load/store from pointer"); + return std::error_code(); +} + /// ParseFunctionBody - Lazily parse the specified function body block. std::error_code BitcodeReader::ParseFunctionBody(Function *F) { if (Stream.EnterSubBlock(bitc::FUNCTION_BLOCK_ID)) @@ -2572,7 +3373,7 @@ std::error_code BitcodeReader::ParseFunctionBody(Function *F) { return EC; break; case bitc::METADATA_ATTACHMENT_ID: - if (std::error_code EC = ParseMetadataAttachment()) + if (std::error_code EC = ParseMetadataAttachment(*F)) return EC; break; case bitc::METADATA_BLOCK_ID: @@ -2731,13 +3532,35 @@ std::error_code BitcodeReader::ParseFunctionBody(Function *F) { InstructionList.push_back(I); break; } - case bitc::FUNC_CODE_INST_INBOUNDS_GEP: - case bitc::FUNC_CODE_INST_GEP: { // GEP: [n x operands] + case bitc::FUNC_CODE_INST_INBOUNDS_GEP_OLD: + case bitc::FUNC_CODE_INST_GEP_OLD: + case bitc::FUNC_CODE_INST_GEP: { // GEP: type, [n x operands] unsigned OpNum = 0; + + Type *Ty; + bool InBounds; + + if (BitCode == bitc::FUNC_CODE_INST_GEP) { + InBounds = Record[OpNum++]; + Ty = getTypeByID(Record[OpNum++]); + } else { + InBounds = BitCode == bitc::FUNC_CODE_INST_INBOUNDS_GEP_OLD; + Ty = nullptr; + } + Value *BasePtr; if (getValueTypePair(Record, OpNum, NextValueNo, BasePtr)) return Error("Invalid record"); + if (!Ty) + Ty = cast<SequentialType>(BasePtr->getType()->getScalarType()) + ->getElementType(); + else if (Ty != + cast<SequentialType>(BasePtr->getType()->getScalarType()) + ->getElementType()) + return Error( + "Explicit gep type does not match pointee type of pointer operand"); + SmallVector<Value*, 16> GEPIdx; while (OpNum != Record.size()) { Value *Op; @@ -2746,9 +3569,10 @@ std::error_code BitcodeReader::ParseFunctionBody(Function *F) { GEPIdx.push_back(Op); } - I = GetElementPtrInst::Create(BasePtr, GEPIdx); + I = GetElementPtrInst::Create(Ty, BasePtr, GEPIdx); + InstructionList.push_back(I); - if (BitCode == bitc::FUNC_CODE_INST_INBOUNDS_GEP) + if (InBounds) cast<GetElementPtrInst>(I)->setIsInBounds(true); break; } @@ -2760,13 +3584,31 @@ std::error_code BitcodeReader::ParseFunctionBody(Function *F) { if (getValueTypePair(Record, OpNum, NextValueNo, Agg)) return Error("Invalid record"); + unsigned RecSize = Record.size(); + if (OpNum == RecSize) + return Error("EXTRACTVAL: Invalid instruction with 0 indices"); + SmallVector<unsigned, 4> EXTRACTVALIdx; - for (unsigned RecSize = Record.size(); - OpNum != RecSize; ++OpNum) { + Type *CurTy = Agg->getType(); + for (; OpNum != RecSize; ++OpNum) { + bool IsArray = CurTy->isArrayTy(); + bool IsStruct = CurTy->isStructTy(); uint64_t Index = Record[OpNum]; + + if (!IsStruct && !IsArray) + return Error("EXTRACTVAL: Invalid type"); if ((unsigned)Index != Index) return Error("Invalid value"); + if (IsStruct && Index >= CurTy->subtypes().size()) + return Error("EXTRACTVAL: Invalid struct index"); + if (IsArray && Index >= CurTy->getArrayNumElements()) + return Error("EXTRACTVAL: Invalid array index"); EXTRACTVALIdx.push_back((unsigned)Index); + + if (IsStruct) + CurTy = CurTy->subtypes()[Index]; + else + CurTy = CurTy->subtypes()[0]; } I = ExtractValueInst::Create(Agg, EXTRACTVALIdx); @@ -2784,15 +3626,36 @@ std::error_code BitcodeReader::ParseFunctionBody(Function *F) { if (getValueTypePair(Record, OpNum, NextValueNo, Val)) return Error("Invalid record"); + unsigned RecSize = Record.size(); + if (OpNum == RecSize) + return Error("INSERTVAL: Invalid instruction with 0 indices"); + SmallVector<unsigned, 4> INSERTVALIdx; - for (unsigned RecSize = Record.size(); - OpNum != RecSize; ++OpNum) { + Type *CurTy = Agg->getType(); + for (; OpNum != RecSize; ++OpNum) { + bool IsArray = CurTy->isArrayTy(); + bool IsStruct = CurTy->isStructTy(); uint64_t Index = Record[OpNum]; + + if (!IsStruct && !IsArray) + return Error("INSERTVAL: Invalid type"); if ((unsigned)Index != Index) return Error("Invalid value"); + if (IsStruct && Index >= CurTy->subtypes().size()) + return Error("INSERTVAL: Invalid struct index"); + if (IsArray && Index >= CurTy->getArrayNumElements()) + return Error("INSERTVAL: Invalid array index"); + INSERTVALIdx.push_back((unsigned)Index); + if (IsStruct) + CurTy = CurTy->subtypes()[Index]; + else + CurTy = CurTy->subtypes()[0]; } + if (CurTy != Val->getType()) + return Error("Inserted value type doesn't match aggregate type"); + I = InsertValueInst::Create(Agg, Val, INSERTVALIdx); InstructionList.push_back(I); break; @@ -2846,6 +3709,8 @@ std::error_code BitcodeReader::ParseFunctionBody(Function *F) { if (getValueTypePair(Record, OpNum, NextValueNo, Vec) || getValueTypePair(Record, OpNum, NextValueNo, Idx)) return Error("Invalid record"); + if (!Vec->getType()->isVectorTy()) + return Error("Invalid type for value"); I = ExtractElementInst::Create(Vec, Idx); InstructionList.push_back(I); break; @@ -2854,8 +3719,11 @@ std::error_code BitcodeReader::ParseFunctionBody(Function *F) { case bitc::FUNC_CODE_INST_INSERTELT: { // INSERTELT: [ty, opval,opval,opval] unsigned OpNum = 0; Value *Vec, *Elt, *Idx; - if (getValueTypePair(Record, OpNum, NextValueNo, Vec) || - popValue(Record, OpNum, NextValueNo, + if (getValueTypePair(Record, OpNum, NextValueNo, Vec)) + return Error("Invalid record"); + if (!Vec->getType()->isVectorTy()) + return Error("Invalid type for value"); + if (popValue(Record, OpNum, NextValueNo, cast<VectorType>(Vec->getType())->getElementType(), Elt) || getValueTypePair(Record, OpNum, NextValueNo, Idx)) return Error("Invalid record"); @@ -2873,6 +3741,8 @@ std::error_code BitcodeReader::ParseFunctionBody(Function *F) { if (getValueTypePair(Record, OpNum, NextValueNo, Mask)) return Error("Invalid record"); + if (!Vec1->getType()->isVectorTy() || !Vec2->getType()->isVectorTy()) + return Error("Invalid type for value"); I = new ShuffleVectorInst(Vec1, Vec2, Mask); InstructionList.push_back(I); break; @@ -3056,24 +3926,33 @@ std::error_code BitcodeReader::ParseFunctionBody(Function *F) { // INVOKE: [attrs, cc, normBB, unwindBB, fnty, op0,op1,op2, ...] if (Record.size() < 4) return Error("Invalid record"); - AttributeSet PAL = getAttributes(Record[0]); - unsigned CCInfo = Record[1]; - BasicBlock *NormalBB = getBasicBlock(Record[2]); - BasicBlock *UnwindBB = getBasicBlock(Record[3]); + unsigned OpNum = 0; + AttributeSet PAL = getAttributes(Record[OpNum++]); + unsigned CCInfo = Record[OpNum++]; + BasicBlock *NormalBB = getBasicBlock(Record[OpNum++]); + BasicBlock *UnwindBB = getBasicBlock(Record[OpNum++]); + + FunctionType *FTy = nullptr; + if (CCInfo >> 13 & 1 && + !(FTy = dyn_cast<FunctionType>(getTypeByID(Record[OpNum++])))) + return Error("Explicit invoke type is not a function type"); - unsigned OpNum = 4; Value *Callee; if (getValueTypePair(Record, OpNum, NextValueNo, Callee)) return Error("Invalid record"); PointerType *CalleeTy = dyn_cast<PointerType>(Callee->getType()); - FunctionType *FTy = !CalleeTy ? nullptr : - dyn_cast<FunctionType>(CalleeTy->getElementType()); - - // Check that the right number of fixed parameters are here. - if (!FTy || !NormalBB || !UnwindBB || - Record.size() < OpNum+FTy->getNumParams()) - return Error("Invalid record"); + if (!CalleeTy) + return Error("Callee is not a pointer"); + if (!FTy) { + FTy = dyn_cast<FunctionType>(CalleeTy->getElementType()); + if (!FTy) + return Error("Callee is not of pointer to function type"); + } else if (CalleeTy->getElementType() != FTy) + return Error("Explicit invoke type does not match pointee type of " + "callee operand"); + if (Record.size() < FTy->getNumParams() + OpNum) + return Error("Insufficient operands to call"); SmallVector<Value*, 16> Ops; for (unsigned i = 0, e = FTy->getNumParams(); i != e; ++i, ++OpNum) { @@ -3098,8 +3977,8 @@ std::error_code BitcodeReader::ParseFunctionBody(Function *F) { I = InvokeInst::Create(Callee, NormalBB, UnwindBB, Ops); InstructionList.push_back(I); - cast<InvokeInst>(I)->setCallingConv( - static_cast<CallingConv::ID>(CCInfo)); + cast<InvokeInst>(I) + ->setCallingConv(static_cast<CallingConv::ID>(~(1U << 13) & CCInfo)); cast<InvokeInst>(I)->setAttributes(PAL); break; } @@ -3187,16 +4066,28 @@ std::error_code BitcodeReader::ParseFunctionBody(Function *F) { case bitc::FUNC_CODE_INST_ALLOCA: { // ALLOCA: [instty, opty, op, align] if (Record.size() != 4) return Error("Invalid record"); - PointerType *Ty = - dyn_cast_or_null<PointerType>(getTypeByID(Record[0])); + uint64_t AlignRecord = Record[3]; + const uint64_t InAllocaMask = uint64_t(1) << 5; + const uint64_t ExplicitTypeMask = uint64_t(1) << 6; + const uint64_t FlagMask = InAllocaMask | ExplicitTypeMask; + bool InAlloca = AlignRecord & InAllocaMask; + Type *Ty = getTypeByID(Record[0]); + if ((AlignRecord & ExplicitTypeMask) == 0) { + auto *PTy = dyn_cast_or_null<PointerType>(Ty); + if (!PTy) + return Error("Old-style alloca with a non-pointer type"); + Ty = PTy->getElementType(); + } Type *OpTy = getTypeByID(Record[1]); Value *Size = getFnValueByID(Record[2], OpTy); - unsigned AlignRecord = Record[3]; - bool InAlloca = AlignRecord & (1 << 5); - unsigned Align = AlignRecord & ((1 << 5) - 1); + unsigned Align; + if (std::error_code EC = + parseAlignmentValue(AlignRecord & ~FlagMask, Align)) { + return EC; + } if (!Ty || !Size) return Error("Invalid record"); - AllocaInst *AI = new AllocaInst(Ty->getElementType(), Size, (1 << Align) >> 1); + AllocaInst *AI = new AllocaInst(Ty, Size, Align); AI->setUsedWithInAlloca(InAlloca); I = AI; InstructionList.push_back(I); @@ -3206,10 +4097,23 @@ std::error_code BitcodeReader::ParseFunctionBody(Function *F) { unsigned OpNum = 0; Value *Op; if (getValueTypePair(Record, OpNum, NextValueNo, Op) || - OpNum+2 != Record.size()) + (OpNum + 2 != Record.size() && OpNum + 3 != Record.size())) return Error("Invalid record"); - I = new LoadInst(Op, "", Record[OpNum+1], (1 << Record[OpNum]) >> 1); + Type *Ty = nullptr; + if (OpNum + 3 == Record.size()) + Ty = getTypeByID(Record[OpNum++]); + if (std::error_code EC = + TypeCheckLoadStoreInst(DiagnosticHandler, Ty, Op->getType())) + return EC; + if (!Ty) + Ty = cast<PointerType>(Op->getType())->getElementType(); + + unsigned Align; + if (std::error_code EC = parseAlignmentValue(Record[OpNum], Align)) + return EC; + I = new LoadInst(Ty, Op, "", Record[OpNum + 1], Align); + InstructionList.push_back(I); break; } @@ -3218,9 +4122,18 @@ std::error_code BitcodeReader::ParseFunctionBody(Function *F) { unsigned OpNum = 0; Value *Op; if (getValueTypePair(Record, OpNum, NextValueNo, Op) || - OpNum+4 != Record.size()) + (OpNum + 4 != Record.size() && OpNum + 5 != Record.size())) return Error("Invalid record"); + Type *Ty = nullptr; + if (OpNum + 5 == Record.size()) + Ty = getTypeByID(Record[OpNum++]); + if (std::error_code EC = + TypeCheckLoadStoreInst(DiagnosticHandler, Ty, Op->getType())) + return EC; + if (!Ty) + Ty = cast<PointerType>(Op->getType())->getElementType(); + AtomicOrdering Ordering = GetDecodedOrdering(Record[OpNum+2]); if (Ordering == NotAtomic || Ordering == Release || Ordering == AcquireRelease) @@ -3229,34 +4142,54 @@ std::error_code BitcodeReader::ParseFunctionBody(Function *F) { return Error("Invalid record"); SynchronizationScope SynchScope = GetDecodedSynchScope(Record[OpNum+3]); - I = new LoadInst(Op, "", Record[OpNum+1], (1 << Record[OpNum]) >> 1, - Ordering, SynchScope); + unsigned Align; + if (std::error_code EC = parseAlignmentValue(Record[OpNum], Align)) + return EC; + I = new LoadInst(Op, "", Record[OpNum+1], Align, Ordering, SynchScope); + InstructionList.push_back(I); break; } - case bitc::FUNC_CODE_INST_STORE: { // STORE2:[ptrty, ptr, val, align, vol] + case bitc::FUNC_CODE_INST_STORE: + case bitc::FUNC_CODE_INST_STORE_OLD: { // STORE2:[ptrty, ptr, val, align, vol] unsigned OpNum = 0; Value *Val, *Ptr; if (getValueTypePair(Record, OpNum, NextValueNo, Ptr) || - popValue(Record, OpNum, NextValueNo, - cast<PointerType>(Ptr->getType())->getElementType(), Val) || - OpNum+2 != Record.size()) + (BitCode == bitc::FUNC_CODE_INST_STORE + ? getValueTypePair(Record, OpNum, NextValueNo, Val) + : popValue(Record, OpNum, NextValueNo, + cast<PointerType>(Ptr->getType())->getElementType(), + Val)) || + OpNum + 2 != Record.size()) return Error("Invalid record"); - I = new StoreInst(Val, Ptr, Record[OpNum+1], (1 << Record[OpNum]) >> 1); + if (std::error_code EC = TypeCheckLoadStoreInst( + DiagnosticHandler, Val->getType(), Ptr->getType())) + return EC; + unsigned Align; + if (std::error_code EC = parseAlignmentValue(Record[OpNum], Align)) + return EC; + I = new StoreInst(Val, Ptr, Record[OpNum+1], Align); InstructionList.push_back(I); break; } - case bitc::FUNC_CODE_INST_STOREATOMIC: { + case bitc::FUNC_CODE_INST_STOREATOMIC: + case bitc::FUNC_CODE_INST_STOREATOMIC_OLD: { // STOREATOMIC: [ptrty, ptr, val, align, vol, ordering, synchscope] unsigned OpNum = 0; Value *Val, *Ptr; if (getValueTypePair(Record, OpNum, NextValueNo, Ptr) || - popValue(Record, OpNum, NextValueNo, - cast<PointerType>(Ptr->getType())->getElementType(), Val) || - OpNum+4 != Record.size()) + (BitCode == bitc::FUNC_CODE_INST_STOREATOMIC + ? getValueTypePair(Record, OpNum, NextValueNo, Val) + : popValue(Record, OpNum, NextValueNo, + cast<PointerType>(Ptr->getType())->getElementType(), + Val)) || + OpNum + 4 != Record.size()) return Error("Invalid record"); + if (std::error_code EC = TypeCheckLoadStoreInst( + DiagnosticHandler, Val->getType(), Ptr->getType())) + return EC; AtomicOrdering Ordering = GetDecodedOrdering(Record[OpNum+2]); if (Ordering == NotAtomic || Ordering == Acquire || Ordering == AcquireRelease) @@ -3265,28 +4198,36 @@ std::error_code BitcodeReader::ParseFunctionBody(Function *F) { if (Ordering != NotAtomic && Record[OpNum] == 0) return Error("Invalid record"); - I = new StoreInst(Val, Ptr, Record[OpNum+1], (1 << Record[OpNum]) >> 1, - Ordering, SynchScope); + unsigned Align; + if (std::error_code EC = parseAlignmentValue(Record[OpNum], Align)) + return EC; + I = new StoreInst(Val, Ptr, Record[OpNum+1], Align, Ordering, SynchScope); InstructionList.push_back(I); break; } + case bitc::FUNC_CODE_INST_CMPXCHG_OLD: case bitc::FUNC_CODE_INST_CMPXCHG: { // CMPXCHG:[ptrty, ptr, cmp, new, vol, successordering, synchscope, // failureordering?, isweak?] unsigned OpNum = 0; Value *Ptr, *Cmp, *New; if (getValueTypePair(Record, OpNum, NextValueNo, Ptr) || - popValue(Record, OpNum, NextValueNo, - cast<PointerType>(Ptr->getType())->getElementType(), Cmp) || - popValue(Record, OpNum, NextValueNo, - cast<PointerType>(Ptr->getType())->getElementType(), New) || - (Record.size() < OpNum + 3 || Record.size() > OpNum + 5)) + (BitCode == bitc::FUNC_CODE_INST_CMPXCHG + ? getValueTypePair(Record, OpNum, NextValueNo, Cmp) + : popValue(Record, OpNum, NextValueNo, + cast<PointerType>(Ptr->getType())->getElementType(), + Cmp)) || + popValue(Record, OpNum, NextValueNo, Cmp->getType(), New) || + Record.size() < OpNum + 3 || Record.size() > OpNum + 5) return Error("Invalid record"); AtomicOrdering SuccessOrdering = GetDecodedOrdering(Record[OpNum+1]); if (SuccessOrdering == NotAtomic || SuccessOrdering == Unordered) return Error("Invalid record"); SynchronizationScope SynchScope = GetDecodedSynchScope(Record[OpNum+2]); + if (std::error_code EC = TypeCheckLoadStoreInst( + DiagnosticHandler, Cmp->getType(), Ptr->getType())) + return EC; AtomicOrdering FailureOrdering; if (Record.size() < 7) FailureOrdering = @@ -3350,19 +4291,31 @@ std::error_code BitcodeReader::ParseFunctionBody(Function *F) { if (Record.size() < 3) return Error("Invalid record"); - AttributeSet PAL = getAttributes(Record[0]); - unsigned CCInfo = Record[1]; + unsigned OpNum = 0; + AttributeSet PAL = getAttributes(Record[OpNum++]); + unsigned CCInfo = Record[OpNum++]; + + FunctionType *FTy = nullptr; + if (CCInfo >> 15 & 1 && + !(FTy = dyn_cast<FunctionType>(getTypeByID(Record[OpNum++])))) + return Error("Explicit call type is not a function type"); - unsigned OpNum = 2; Value *Callee; if (getValueTypePair(Record, OpNum, NextValueNo, Callee)) return Error("Invalid record"); PointerType *OpTy = dyn_cast<PointerType>(Callee->getType()); - FunctionType *FTy = nullptr; - if (OpTy) FTy = dyn_cast<FunctionType>(OpTy->getElementType()); - if (!FTy || Record.size() < FTy->getNumParams()+OpNum) - return Error("Invalid record"); + if (!OpTy) + return Error("Callee is not a pointer type"); + if (!FTy) { + FTy = dyn_cast<FunctionType>(OpTy->getElementType()); + if (!FTy) + return Error("Callee is not of pointer to function type"); + } else if (OpTy->getElementType() != FTy) + return Error("Explicit call type does not match pointee type of " + "callee operand"); + if (Record.size() < FTy->getNumParams() + OpNum) + return Error("Insufficient operands to call"); SmallVector<Value*, 16> Args; // Read the fixed params. @@ -3389,7 +4342,7 @@ std::error_code BitcodeReader::ParseFunctionBody(Function *F) { } } - I = CallInst::Create(Callee, Args); + I = CallInst::Create(FTy, Callee, Args); InstructionList.push_back(I); cast<CallInst>(I)->setCallingConv( static_cast<CallingConv::ID>((~(1U << 14) & CCInfo) >> 1)); @@ -3483,6 +4436,9 @@ std::error_code BitcodeReader::FindFunctionInStream( void BitcodeReader::releaseBuffer() { Buffer.release(); } std::error_code BitcodeReader::materialize(GlobalValue *GV) { + if (std::error_code EC = materializeMetadata()) + return EC; + Function *F = dyn_cast<Function>(GV); // If it's not a function or is already material, ignore the request. if (!F || !F->isMaterializable()) @@ -3503,6 +4459,9 @@ std::error_code BitcodeReader::materialize(GlobalValue *GV) { return EC; F->setIsMaterializable(false); + if (StripDebugInfo) + stripDebugInfo(*F); + // Upgrade any old intrinsic calls in the function. for (UpgradedIntrinsicMap::iterator I = UpgradedIntrinsics.begin(), E = UpgradedIntrinsics.end(); I != E; ++I) { @@ -3533,7 +4492,7 @@ bool BitcodeReader::isDematerializable(const GlobalValue *GV) const { return DeferredFunctionInfo.count(const_cast<Function*>(F)); } -void BitcodeReader::Dematerialize(GlobalValue *GV) { +void BitcodeReader::dematerialize(GlobalValue *GV) { Function *F = dyn_cast<Function>(GV); // If this function isn't dematerializable, this is a noop. if (!F || !isDematerializable(F)) @@ -3546,10 +4505,13 @@ void BitcodeReader::Dematerialize(GlobalValue *GV) { F->setIsMaterializable(true); } -std::error_code BitcodeReader::MaterializeModule(Module *M) { +std::error_code BitcodeReader::materializeModule(Module *M) { assert(M == TheModule && "Can only Materialize the Module this BitcodeReader is attached to."); + if (std::error_code EC = materializeMetadata()) + return EC; + // Promise to materialize all forward references. WillMaterializeAllForwardRefs = true; @@ -3690,7 +4652,8 @@ const std::error_category &llvm::BitcodeErrorCategory() { static ErrorOr<Module *> getLazyBitcodeModuleImpl(std::unique_ptr<MemoryBuffer> &&Buffer, LLVMContext &Context, bool WillMaterializeAll, - DiagnosticHandlerFunction DiagnosticHandler) { + DiagnosticHandlerFunction DiagnosticHandler, + bool ShouldLazyLoadMetadata = false) { Module *M = new Module(Buffer->getBufferIdentifier(), Context); BitcodeReader *R = new BitcodeReader(Buffer.get(), Context, DiagnosticHandler); @@ -3702,7 +4665,8 @@ getLazyBitcodeModuleImpl(std::unique_ptr<MemoryBuffer> &&Buffer, return EC; }; - if (std::error_code EC = R->ParseBitcodeInto(M)) + // Delay parsing Metadata if ShouldLazyLoadMetadata is true. + if (std::error_code EC = R->ParseBitcodeInto(M, ShouldLazyLoadMetadata)) return cleanupOnError(EC); if (!WillMaterializeAll) @@ -3717,9 +4681,10 @@ getLazyBitcodeModuleImpl(std::unique_ptr<MemoryBuffer> &&Buffer, ErrorOr<Module *> llvm::getLazyBitcodeModule(std::unique_ptr<MemoryBuffer> &&Buffer, LLVMContext &Context, - DiagnosticHandlerFunction DiagnosticHandler) { + DiagnosticHandlerFunction DiagnosticHandler, + bool ShouldLazyLoadMetadata) { return getLazyBitcodeModuleImpl(std::move(Buffer), Context, false, - DiagnosticHandler); + DiagnosticHandler, ShouldLazyLoadMetadata); } ErrorOr<std::unique_ptr<Module>> diff --git a/lib/Bitcode/Reader/BitcodeReader.h b/lib/Bitcode/Reader/BitcodeReader.h deleted file mode 100644 index 5090be48450e..000000000000 --- a/lib/Bitcode/Reader/BitcodeReader.h +++ /dev/null @@ -1,365 +0,0 @@ -//===- BitcodeReader.h - Internal BitcodeReader impl ------------*- C++ -*-===// -// -// The LLVM Compiler Infrastructure -// -// This file is distributed under the University of Illinois Open Source -// License. See LICENSE.TXT for details. -// -//===----------------------------------------------------------------------===// -// -// This header defines the BitcodeReader class. -// -//===----------------------------------------------------------------------===// - -#ifndef LLVM_LIB_BITCODE_READER_BITCODEREADER_H -#define LLVM_LIB_BITCODE_READER_BITCODEREADER_H - -#include "llvm/ADT/DenseMap.h" -#include "llvm/Bitcode/BitstreamReader.h" -#include "llvm/Bitcode/LLVMBitCodes.h" -#include "llvm/IR/Attributes.h" -#include "llvm/IR/GVMaterializer.h" -#include "llvm/IR/Metadata.h" -#include "llvm/IR/OperandTraits.h" -#include "llvm/IR/TrackingMDRef.h" -#include "llvm/IR/Type.h" -#include "llvm/IR/ValueHandle.h" -#include <deque> -#include <system_error> -#include <vector> - -namespace llvm { - class Comdat; - class MemoryBuffer; - class LLVMContext; - -//===----------------------------------------------------------------------===// -// BitcodeReaderValueList Class -//===----------------------------------------------------------------------===// - -class BitcodeReaderValueList { - std::vector<WeakVH> ValuePtrs; - - /// ResolveConstants - As we resolve forward-referenced constants, we add - /// information about them to this vector. This allows us to resolve them in - /// bulk instead of resolving each reference at a time. See the code in - /// ResolveConstantForwardRefs for more information about this. - /// - /// The key of this vector is the placeholder constant, the value is the slot - /// number that holds the resolved value. - typedef std::vector<std::pair<Constant*, unsigned> > ResolveConstantsTy; - ResolveConstantsTy ResolveConstants; - LLVMContext &Context; -public: - BitcodeReaderValueList(LLVMContext &C) : Context(C) {} - ~BitcodeReaderValueList() { - assert(ResolveConstants.empty() && "Constants not resolved?"); - } - - // vector compatibility methods - unsigned size() const { return ValuePtrs.size(); } - void resize(unsigned N) { ValuePtrs.resize(N); } - void push_back(Value *V) { - ValuePtrs.push_back(V); - } - - void clear() { - assert(ResolveConstants.empty() && "Constants not resolved?"); - ValuePtrs.clear(); - } - - Value *operator[](unsigned i) const { - assert(i < ValuePtrs.size()); - return ValuePtrs[i]; - } - - Value *back() const { return ValuePtrs.back(); } - void pop_back() { ValuePtrs.pop_back(); } - bool empty() const { return ValuePtrs.empty(); } - void shrinkTo(unsigned N) { - assert(N <= size() && "Invalid shrinkTo request!"); - ValuePtrs.resize(N); - } - - Constant *getConstantFwdRef(unsigned Idx, Type *Ty); - Value *getValueFwdRef(unsigned Idx, Type *Ty); - - void AssignValue(Value *V, unsigned Idx); - - /// ResolveConstantForwardRefs - Once all constants are read, this method bulk - /// resolves any forward references. - void ResolveConstantForwardRefs(); -}; - - -//===----------------------------------------------------------------------===// -// BitcodeReaderMDValueList Class -//===----------------------------------------------------------------------===// - -class BitcodeReaderMDValueList { - unsigned NumFwdRefs; - bool AnyFwdRefs; - unsigned MinFwdRef; - unsigned MaxFwdRef; - std::vector<TrackingMDRef> MDValuePtrs; - - LLVMContext &Context; -public: - BitcodeReaderMDValueList(LLVMContext &C) - : NumFwdRefs(0), AnyFwdRefs(false), Context(C) {} - - // vector compatibility methods - unsigned size() const { return MDValuePtrs.size(); } - void resize(unsigned N) { MDValuePtrs.resize(N); } - void push_back(Metadata *MD) { MDValuePtrs.emplace_back(MD); } - void clear() { MDValuePtrs.clear(); } - Metadata *back() const { return MDValuePtrs.back(); } - void pop_back() { MDValuePtrs.pop_back(); } - bool empty() const { return MDValuePtrs.empty(); } - - Metadata *operator[](unsigned i) const { - assert(i < MDValuePtrs.size()); - return MDValuePtrs[i]; - } - - void shrinkTo(unsigned N) { - assert(N <= size() && "Invalid shrinkTo request!"); - MDValuePtrs.resize(N); - } - - Metadata *getValueFwdRef(unsigned Idx); - void AssignValue(Metadata *MD, unsigned Idx); - void tryToResolveCycles(); -}; - -class BitcodeReader : public GVMaterializer { - LLVMContext &Context; - DiagnosticHandlerFunction DiagnosticHandler; - Module *TheModule; - std::unique_ptr<MemoryBuffer> Buffer; - std::unique_ptr<BitstreamReader> StreamFile; - BitstreamCursor Stream; - DataStreamer *LazyStreamer; - uint64_t NextUnreadBit; - bool SeenValueSymbolTable; - - std::vector<Type*> TypeList; - BitcodeReaderValueList ValueList; - BitcodeReaderMDValueList MDValueList; - std::vector<Comdat *> ComdatList; - SmallVector<Instruction *, 64> InstructionList; - - std::vector<std::pair<GlobalVariable*, unsigned> > GlobalInits; - std::vector<std::pair<GlobalAlias*, unsigned> > AliasInits; - std::vector<std::pair<Function*, unsigned> > FunctionPrefixes; - std::vector<std::pair<Function*, unsigned> > FunctionPrologues; - - SmallVector<Instruction*, 64> InstsWithTBAATag; - - /// MAttributes - The set of attributes by index. Index zero in the - /// file is for null, and is thus not represented here. As such all indices - /// are off by one. - std::vector<AttributeSet> MAttributes; - - /// \brief The set of attribute groups. - std::map<unsigned, AttributeSet> MAttributeGroups; - - /// FunctionBBs - While parsing a function body, this is a list of the basic - /// blocks for the function. - std::vector<BasicBlock*> FunctionBBs; - - // When reading the module header, this list is populated with functions that - // have bodies later in the file. - std::vector<Function*> FunctionsWithBodies; - - // When intrinsic functions are encountered which require upgrading they are - // stored here with their replacement function. - typedef std::vector<std::pair<Function*, Function*> > UpgradedIntrinsicMap; - UpgradedIntrinsicMap UpgradedIntrinsics; - - // Map the bitcode's custom MDKind ID to the Module's MDKind ID. - DenseMap<unsigned, unsigned> MDKindMap; - - // Several operations happen after the module header has been read, but - // before function bodies are processed. This keeps track of whether - // we've done this yet. - bool SeenFirstFunctionBody; - - /// DeferredFunctionInfo - When function bodies are initially scanned, this - /// map contains info about where to find deferred function body in the - /// stream. - DenseMap<Function*, uint64_t> DeferredFunctionInfo; - - /// These are basic blocks forward-referenced by block addresses. They are - /// inserted lazily into functions when they're loaded. The basic block ID is - /// its index into the vector. - DenseMap<Function *, std::vector<BasicBlock *>> BasicBlockFwdRefs; - std::deque<Function *> BasicBlockFwdRefQueue; - - /// UseRelativeIDs - Indicates that we are using a new encoding for - /// instruction operands where most operands in the current - /// FUNCTION_BLOCK are encoded relative to the instruction number, - /// for a more compact encoding. Some instruction operands are not - /// relative to the instruction ID: basic block numbers, and types. - /// Once the old style function blocks have been phased out, we would - /// not need this flag. - bool UseRelativeIDs; - - /// True if all functions will be materialized, negating the need to process - /// (e.g.) blockaddress forward references. - bool WillMaterializeAllForwardRefs; - - /// Functions that have block addresses taken. This is usually empty. - SmallPtrSet<const Function *, 4> BlockAddressesTaken; - -public: - std::error_code Error(BitcodeError E, const Twine &Message); - std::error_code Error(BitcodeError E); - std::error_code Error(const Twine &Message); - - explicit BitcodeReader(MemoryBuffer *buffer, LLVMContext &C, - DiagnosticHandlerFunction DiagnosticHandler); - explicit BitcodeReader(DataStreamer *streamer, LLVMContext &C, - DiagnosticHandlerFunction DiagnosticHandler); - ~BitcodeReader() { FreeState(); } - - std::error_code materializeForwardReferencedFunctions(); - - void FreeState(); - - void releaseBuffer(); - - bool isDematerializable(const GlobalValue *GV) const override; - std::error_code materialize(GlobalValue *GV) override; - std::error_code MaterializeModule(Module *M) override; - std::vector<StructType *> getIdentifiedStructTypes() const override; - void Dematerialize(GlobalValue *GV) override; - - /// @brief Main interface to parsing a bitcode buffer. - /// @returns true if an error occurred. - std::error_code ParseBitcodeInto(Module *M); - - /// @brief Cheap mechanism to just extract module triple - /// @returns true if an error occurred. - ErrorOr<std::string> parseTriple(); - - static uint64_t decodeSignRotatedValue(uint64_t V); - -private: - std::vector<StructType *> IdentifiedStructTypes; - StructType *createIdentifiedStructType(LLVMContext &Context, StringRef Name); - StructType *createIdentifiedStructType(LLVMContext &Context); - - Type *getTypeByID(unsigned ID); - Value *getFnValueByID(unsigned ID, Type *Ty) { - if (Ty && Ty->isMetadataTy()) - return MetadataAsValue::get(Ty->getContext(), getFnMetadataByID(ID)); - return ValueList.getValueFwdRef(ID, Ty); - } - Metadata *getFnMetadataByID(unsigned ID) { - return MDValueList.getValueFwdRef(ID); - } - BasicBlock *getBasicBlock(unsigned ID) const { - if (ID >= FunctionBBs.size()) return nullptr; // Invalid ID - return FunctionBBs[ID]; - } - AttributeSet getAttributes(unsigned i) const { - if (i-1 < MAttributes.size()) - return MAttributes[i-1]; - return AttributeSet(); - } - - /// getValueTypePair - Read a value/type pair out of the specified record from - /// slot 'Slot'. Increment Slot past the number of slots used in the record. - /// Return true on failure. - bool getValueTypePair(SmallVectorImpl<uint64_t> &Record, unsigned &Slot, - unsigned InstNum, Value *&ResVal) { - if (Slot == Record.size()) return true; - unsigned ValNo = (unsigned)Record[Slot++]; - // Adjust the ValNo, if it was encoded relative to the InstNum. - if (UseRelativeIDs) - ValNo = InstNum - ValNo; - if (ValNo < InstNum) { - // If this is not a forward reference, just return the value we already - // have. - ResVal = getFnValueByID(ValNo, nullptr); - return ResVal == nullptr; - } else if (Slot == Record.size()) { - return true; - } - - unsigned TypeNo = (unsigned)Record[Slot++]; - ResVal = getFnValueByID(ValNo, getTypeByID(TypeNo)); - return ResVal == nullptr; - } - - /// popValue - Read a value out of the specified record from slot 'Slot'. - /// Increment Slot past the number of slots used by the value in the record. - /// Return true if there is an error. - bool popValue(SmallVectorImpl<uint64_t> &Record, unsigned &Slot, - unsigned InstNum, Type *Ty, Value *&ResVal) { - if (getValue(Record, Slot, InstNum, Ty, ResVal)) - return true; - // All values currently take a single record slot. - ++Slot; - return false; - } - - /// getValue -- Like popValue, but does not increment the Slot number. - bool getValue(SmallVectorImpl<uint64_t> &Record, unsigned Slot, - unsigned InstNum, Type *Ty, Value *&ResVal) { - ResVal = getValue(Record, Slot, InstNum, Ty); - return ResVal == nullptr; - } - - /// getValue -- Version of getValue that returns ResVal directly, - /// or 0 if there is an error. - Value *getValue(SmallVectorImpl<uint64_t> &Record, unsigned Slot, - unsigned InstNum, Type *Ty) { - if (Slot == Record.size()) return nullptr; - unsigned ValNo = (unsigned)Record[Slot]; - // Adjust the ValNo, if it was encoded relative to the InstNum. - if (UseRelativeIDs) - ValNo = InstNum - ValNo; - return getFnValueByID(ValNo, Ty); - } - - /// getValueSigned -- Like getValue, but decodes signed VBRs. - Value *getValueSigned(SmallVectorImpl<uint64_t> &Record, unsigned Slot, - unsigned InstNum, Type *Ty) { - if (Slot == Record.size()) return nullptr; - unsigned ValNo = (unsigned)decodeSignRotatedValue(Record[Slot]); - // Adjust the ValNo, if it was encoded relative to the InstNum. - if (UseRelativeIDs) - ValNo = InstNum - ValNo; - return getFnValueByID(ValNo, Ty); - } - - std::error_code ParseAttrKind(uint64_t Code, Attribute::AttrKind *Kind); - std::error_code ParseModule(bool Resume); - std::error_code ParseAttributeBlock(); - std::error_code ParseAttributeGroupBlock(); - std::error_code ParseTypeTable(); - std::error_code ParseTypeTableBody(); - - std::error_code ParseValueSymbolTable(); - std::error_code ParseConstants(); - std::error_code RememberAndSkipFunctionBody(); - std::error_code ParseFunctionBody(Function *F); - std::error_code GlobalCleanup(); - std::error_code ResolveGlobalAndAliasInits(); - std::error_code ParseMetadata(); - std::error_code ParseMetadataAttachment(); - ErrorOr<std::string> parseModuleTriple(); - std::error_code ParseUseLists(); - std::error_code InitStream(); - std::error_code InitStreamFromBuffer(); - std::error_code InitLazyStream(); - std::error_code FindFunctionInStream( - Function *F, - DenseMap<Function *, uint64_t>::iterator DeferredFunctionInfoIterator); -}; - -} // End llvm namespace - -#endif diff --git a/lib/Bitcode/Reader/BitstreamReader.cpp b/lib/Bitcode/Reader/BitstreamReader.cpp index 5e3232e53135..a103fbdf4a93 100644 --- a/lib/Bitcode/Reader/BitstreamReader.cpp +++ b/lib/Bitcode/Reader/BitstreamReader.cpp @@ -39,15 +39,16 @@ bool BitstreamCursor::EnterSubBlock(unsigned BlockID, unsigned *NumWordsP) { // Get the codesize of this block. CurCodeSize = ReadVBR(bitc::CodeLenWidth); + // We can't read more than MaxChunkSize at a time + if (CurCodeSize > MaxChunkSize) + return true; + SkipToFourByteBoundary(); unsigned NumWords = Read(bitc::BlockSizeWidth); if (NumWordsP) *NumWordsP = NumWords; // Validate that this block is sane. - if (CurCodeSize == 0 || AtEndOfStream()) - return true; - - return false; + return CurCodeSize == 0 || AtEndOfStream(); } static uint64_t readAbbreviatedField(BitstreamCursor &Cursor, @@ -60,8 +61,10 @@ static uint64_t readAbbreviatedField(BitstreamCursor &Cursor, case BitCodeAbbrevOp::Blob: llvm_unreachable("Should not reach here"); case BitCodeAbbrevOp::Fixed: + assert((unsigned)Op.getEncodingData() <= Cursor.MaxChunkSize); return Cursor.Read((unsigned)Op.getEncodingData()); case BitCodeAbbrevOp::VBR: + assert((unsigned)Op.getEncodingData() <= Cursor.MaxChunkSize); return Cursor.ReadVBR64((unsigned)Op.getEncodingData()); case BitCodeAbbrevOp::Char6: return BitCodeAbbrevOp::DecodeChar6(Cursor.Read(6)); @@ -79,9 +82,11 @@ static void skipAbbreviatedField(BitstreamCursor &Cursor, case BitCodeAbbrevOp::Blob: llvm_unreachable("Should not reach here"); case BitCodeAbbrevOp::Fixed: + assert((unsigned)Op.getEncodingData() <= Cursor.MaxChunkSize); Cursor.Read((unsigned)Op.getEncodingData()); break; case BitCodeAbbrevOp::VBR: + assert((unsigned)Op.getEncodingData() <= Cursor.MaxChunkSize); Cursor.ReadVBR64((unsigned)Op.getEncodingData()); break; case BitCodeAbbrevOp::Char6: @@ -170,8 +175,12 @@ unsigned BitstreamCursor::readRecord(unsigned AbbrevID, unsigned Code; if (CodeOp.isLiteral()) Code = CodeOp.getLiteralValue(); - else + else { + if (CodeOp.getEncoding() == BitCodeAbbrevOp::Array || + CodeOp.getEncoding() == BitCodeAbbrevOp::Blob) + report_fatal_error("Abbreviation starts with an Array or a Blob"); Code = readAbbreviatedField(*this, CodeOp); + } for (unsigned i = 1, e = Abbv->getNumOperandInfos(); i != e; ++i) { const BitCodeAbbrevOp &Op = Abbv->getOperandInfo(i); @@ -191,8 +200,15 @@ unsigned BitstreamCursor::readRecord(unsigned AbbrevID, unsigned NumElts = ReadVBR(6); // Get the element encoding. - assert(i+2 == e && "array op not second to last?"); + if (i + 2 != e) + report_fatal_error("Array op not second to last"); const BitCodeAbbrevOp &EltEnc = Abbv->getOperandInfo(++i); + if (!EltEnc.isEncoding()) + report_fatal_error( + "Array element type has to be an encoding of a type"); + if (EltEnc.getEncoding() == BitCodeAbbrevOp::Array || + EltEnc.getEncoding() == BitCodeAbbrevOp::Blob) + report_fatal_error("Array element type can't be an Array or a Blob"); // Read all the elements. for (; NumElts; --NumElts) @@ -241,7 +257,7 @@ void BitstreamCursor::ReadAbbrevRecord() { BitCodeAbbrev *Abbv = new BitCodeAbbrev(); unsigned NumOpInfo = ReadVBR(5); for (unsigned i = 0; i != NumOpInfo; ++i) { - bool IsLiteral = Read(1) ? true : false; + bool IsLiteral = Read(1); if (IsLiteral) { Abbv->Add(BitCodeAbbrevOp(ReadVBR64(8))); continue; @@ -249,7 +265,7 @@ void BitstreamCursor::ReadAbbrevRecord() { BitCodeAbbrevOp::Encoding E = (BitCodeAbbrevOp::Encoding)Read(3); if (BitCodeAbbrevOp::hasEncodingData(E)) { - unsigned Data = ReadVBR64(5); + uint64_t Data = ReadVBR64(5); // As a special case, handle fixed(0) (i.e., a fixed field with zero bits) // and vbr(0) as a literal zero. This is decoded the same way, and avoids @@ -260,10 +276,18 @@ void BitstreamCursor::ReadAbbrevRecord() { continue; } + if ((E == BitCodeAbbrevOp::Fixed || E == BitCodeAbbrevOp::VBR) && + Data > MaxChunkSize) + report_fatal_error( + "Fixed or VBR abbrev record with size > MaxChunkData"); + Abbv->Add(BitCodeAbbrevOp(E, Data)); } else Abbv->Add(BitCodeAbbrevOp(E)); } + + if (Abbv->getNumOperandInfos() == 0) + report_fatal_error("Abbrev record with no operands"); CurAbbrevs.push_back(Abbv); } diff --git a/lib/Bitcode/Reader/CMakeLists.txt b/lib/Bitcode/Reader/CMakeLists.txt index f614c9fd4a03..62954f26e208 100644 --- a/lib/Bitcode/Reader/CMakeLists.txt +++ b/lib/Bitcode/Reader/CMakeLists.txt @@ -2,6 +2,9 @@ add_llvm_library(LLVMBitReader BitReader.cpp BitcodeReader.cpp BitstreamReader.cpp + + ADDITIONAL_HEADER_DIRS + ${LLVM_MAIN_INCLUDE_DIR}/llvm/Bitcode ) add_dependencies(LLVMBitReader intrinsics_gen) diff --git a/lib/Bitcode/Writer/BitcodeWriter.cpp b/lib/Bitcode/Writer/BitcodeWriter.cpp index a96e866ed2c7..97caefb4c494 100644 --- a/lib/Bitcode/Writer/BitcodeWriter.cpp +++ b/lib/Bitcode/Writer/BitcodeWriter.cpp @@ -17,6 +17,7 @@ #include "llvm/Bitcode/BitstreamWriter.h" #include "llvm/Bitcode/LLVMBitCodes.h" #include "llvm/IR/Constants.h" +#include "llvm/IR/DebugInfoMetadata.h" #include "llvm/IR/DerivedTypes.h" #include "llvm/IR/InlineAsm.h" #include "llvm/IR/Instructions.h" @@ -55,7 +56,8 @@ enum { FUNCTION_INST_CAST_ABBREV, FUNCTION_INST_RET_VOID_ABBREV, FUNCTION_INST_RET_VAL_ABBREV, - FUNCTION_INST_UNREACHABLE_ABBREV + FUNCTION_INST_UNREACHABLE_ABBREV, + FUNCTION_INST_GEP_ABBREV, }; static unsigned GetEncodedCastOpcode(unsigned Opcode) { @@ -164,6 +166,8 @@ static uint64_t getAttrKindEncoding(Attribute::AttrKind Kind) { return bitc::ATTR_KIND_BUILTIN; case Attribute::ByVal: return bitc::ATTR_KIND_BY_VAL; + case Attribute::Convergent: + return bitc::ATTR_KIND_CONVERGENT; case Attribute::InAlloca: return bitc::ATTR_KIND_IN_ALLOCA; case Attribute::Cold: @@ -198,6 +202,8 @@ static uint64_t getAttrKindEncoding(Attribute::AttrKind Kind) { return bitc::ATTR_KIND_NON_NULL; case Attribute::Dereferenceable: return bitc::ATTR_KIND_DEREFERENCEABLE; + case Attribute::DereferenceableOrNull: + return bitc::ATTR_KIND_DEREFERENCEABLE_OR_NULL; case Attribute::NoRedZone: return bitc::ATTR_KIND_NO_RED_ZONE; case Attribute::NoReturn: @@ -322,7 +328,7 @@ static void WriteTypeTable(const ValueEnumerator &VE, BitstreamWriter &Stream) { Stream.EnterSubblock(bitc::TYPE_BLOCK_ID_NEW, 4 /*count from # abbrevs */); SmallVector<uint64_t, 64> TypeVals; - uint64_t NumBits = Log2_32_Ceil(VE.getTypes().size()+1); + uint64_t NumBits = VE.computeBitsRequiredForTypeIndicies(); // Abbrev for TYPE_CODE_POINTER. BitCodeAbbrev *Abbv = new BitCodeAbbrev(); @@ -480,13 +486,13 @@ static unsigned getEncodedLinkage(const GlobalValue &GV) { case GlobalValue::ExternalLinkage: return 0; case GlobalValue::WeakAnyLinkage: - return 1; + return 16; case GlobalValue::AppendingLinkage: return 2; case GlobalValue::InternalLinkage: return 3; case GlobalValue::LinkOnceAnyLinkage: - return 4; + return 18; case GlobalValue::ExternalWeakLinkage: return 7; case GlobalValue::CommonLinkage: @@ -494,9 +500,9 @@ static unsigned getEncodedLinkage(const GlobalValue &GV) { case GlobalValue::PrivateLinkage: return 9; case GlobalValue::WeakODRLinkage: - return 10; + return 17; case GlobalValue::LinkOnceODRLinkage: - return 11; + return 19; case GlobalValue::AvailableExternallyLinkage: return 12; } @@ -549,11 +555,13 @@ static unsigned getEncodedComdatSelectionKind(const Comdat &C) { } static void writeComdats(const ValueEnumerator &VE, BitstreamWriter &Stream) { - SmallVector<uint8_t, 64> Vals; + SmallVector<uint16_t, 64> Vals; for (const Comdat *C : VE.getComdats()) { // COMDAT: [selection_kind, name] Vals.push_back(getEncodedComdatSelectionKind(*C)); - Vals.push_back(C->getName().size()); + size_t Size = C->getName().size(); + assert(isUInt<16>(Size)); + Vals.push_back(Size); for (char Chr : C->getName()) Vals.push_back((unsigned char)Chr); Stream.EmitRecord(bitc::MODULE_CODE_COMDAT, Vals, /*AbbrevToUse=*/0); @@ -584,7 +592,7 @@ static void WriteModuleInfo(const Module *M, const ValueEnumerator &VE, unsigned MaxGlobalType = 0; for (const GlobalValue &GV : M->globals()) { MaxAlignment = std::max(MaxAlignment, GV.getAlignment()); - MaxGlobalType = std::max(MaxGlobalType, VE.getTypeID(GV.getType())); + MaxGlobalType = std::max(MaxGlobalType, VE.getTypeID(GV.getValueType())); if (GV.hasSection()) { // Give section names unique ID's. unsigned &Entry = SectionMap[GV.getSection()]; @@ -625,10 +633,12 @@ static void WriteModuleInfo(const Module *M, const ValueEnumerator &VE, Abbv->Add(BitCodeAbbrevOp(bitc::MODULE_CODE_GLOBALVAR)); Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::Fixed, Log2_32_Ceil(MaxGlobalType+1))); - Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::Fixed, 1)); // Constant. - Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::VBR, 6)); // Initializer. - Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::Fixed, 4)); // Linkage. - if (MaxAlignment == 0) // Alignment. + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::VBR, 6)); // AddrSpace << 2 + //| explicitType << 1 + //| constant + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::VBR, 6)); // Initializer. + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::Fixed, 5)); // Linkage. + if (MaxAlignment == 0) // Alignment. Abbv->Add(BitCodeAbbrevOp(0)); else { unsigned MaxEncAlignment = Log2_32(MaxAlignment)+1; @@ -651,9 +661,10 @@ static void WriteModuleInfo(const Module *M, const ValueEnumerator &VE, // GLOBALVAR: [type, isconst, initid, // linkage, alignment, section, visibility, threadlocal, - // unnamed_addr, externally_initialized, dllstorageclass] - Vals.push_back(VE.getTypeID(GV.getType())); - Vals.push_back(GV.isConstant()); + // unnamed_addr, externally_initialized, dllstorageclass, + // comdat] + Vals.push_back(VE.getTypeID(GV.getValueType())); + Vals.push_back(GV.getType()->getAddressSpace() << 2 | 2 | GV.isConstant()); Vals.push_back(GV.isDeclaration() ? 0 : (VE.getValueID(GV.getInitializer()) + 1)); Vals.push_back(getEncodedLinkage(GV)); @@ -683,7 +694,7 @@ static void WriteModuleInfo(const Module *M, const ValueEnumerator &VE, // FUNCTION: [type, callingconv, isproto, linkage, paramattrs, alignment, // section, visibility, gc, unnamed_addr, prologuedata, // dllstorageclass, comdat, prefixdata] - Vals.push_back(VE.getTypeID(F.getType())); + Vals.push_back(VE.getTypeID(F.getFunctionType())); Vals.push_back(F.getCallingConv()); Vals.push_back(F.isDeclaration()); Vals.push_back(getEncodedLinkage(F)); @@ -760,26 +771,22 @@ static void WriteValueAsMetadata(const ValueAsMetadata *MD, Record.clear(); } -static void WriteMDNode(const MDNode *N, - const ValueEnumerator &VE, - BitstreamWriter &Stream, - SmallVectorImpl<uint64_t> &Record) { +static void WriteMDTuple(const MDTuple *N, const ValueEnumerator &VE, + BitstreamWriter &Stream, + SmallVectorImpl<uint64_t> &Record, unsigned Abbrev) { for (unsigned i = 0, e = N->getNumOperands(); i != e; ++i) { Metadata *MD = N->getOperand(i); - if (!MD) { - Record.push_back(0); - continue; - } - assert(!isa<LocalAsMetadata>(MD) && "Unexpected function-local metadata"); - Record.push_back(VE.getMetadataID(MD) + 1); + assert(!(MD && isa<LocalAsMetadata>(MD)) && + "Unexpected function-local metadata"); + Record.push_back(VE.getMetadataOrNullID(MD)); } Stream.EmitRecord(N->isDistinct() ? bitc::METADATA_DISTINCT_NODE : bitc::METADATA_NODE, - Record); + Record, Abbrev); Record.clear(); } -static void WriteMDLocation(const MDLocation *N, const ValueEnumerator &VE, +static void WriteDILocation(const DILocation *N, const ValueEnumerator &VE, BitstreamWriter &Stream, SmallVectorImpl<uint64_t> &Record, unsigned Abbrev) { @@ -787,17 +794,355 @@ static void WriteMDLocation(const MDLocation *N, const ValueEnumerator &VE, Record.push_back(N->getLine()); Record.push_back(N->getColumn()); Record.push_back(VE.getMetadataID(N->getScope())); - - // Always emit the inlined-at location, even though it's optional. - if (Metadata *InlinedAt = N->getInlinedAt()) - Record.push_back(VE.getMetadataID(InlinedAt) + 1); - else - Record.push_back(0); + Record.push_back(VE.getMetadataOrNullID(N->getInlinedAt())); Stream.EmitRecord(bitc::METADATA_LOCATION, Record, Abbrev); Record.clear(); } +static void WriteGenericDINode(const GenericDINode *N, + const ValueEnumerator &VE, + BitstreamWriter &Stream, + SmallVectorImpl<uint64_t> &Record, + unsigned Abbrev) { + Record.push_back(N->isDistinct()); + Record.push_back(N->getTag()); + Record.push_back(0); // Per-tag version field; unused for now. + + for (auto &I : N->operands()) + Record.push_back(VE.getMetadataOrNullID(I)); + + Stream.EmitRecord(bitc::METADATA_GENERIC_DEBUG, Record, Abbrev); + Record.clear(); +} + +static uint64_t rotateSign(int64_t I) { + uint64_t U = I; + return I < 0 ? ~(U << 1) : U << 1; +} + +static void WriteDISubrange(const DISubrange *N, const ValueEnumerator &, + BitstreamWriter &Stream, + SmallVectorImpl<uint64_t> &Record, + unsigned Abbrev) { + Record.push_back(N->isDistinct()); + Record.push_back(N->getCount()); + Record.push_back(rotateSign(N->getLowerBound())); + + Stream.EmitRecord(bitc::METADATA_SUBRANGE, Record, Abbrev); + Record.clear(); +} + +static void WriteDIEnumerator(const DIEnumerator *N, const ValueEnumerator &VE, + BitstreamWriter &Stream, + SmallVectorImpl<uint64_t> &Record, + unsigned Abbrev) { + Record.push_back(N->isDistinct()); + Record.push_back(rotateSign(N->getValue())); + Record.push_back(VE.getMetadataOrNullID(N->getRawName())); + + Stream.EmitRecord(bitc::METADATA_ENUMERATOR, Record, Abbrev); + Record.clear(); +} + +static void WriteDIBasicType(const DIBasicType *N, const ValueEnumerator &VE, + BitstreamWriter &Stream, + SmallVectorImpl<uint64_t> &Record, + unsigned Abbrev) { + Record.push_back(N->isDistinct()); + Record.push_back(N->getTag()); + Record.push_back(VE.getMetadataOrNullID(N->getRawName())); + Record.push_back(N->getSizeInBits()); + Record.push_back(N->getAlignInBits()); + Record.push_back(N->getEncoding()); + + Stream.EmitRecord(bitc::METADATA_BASIC_TYPE, Record, Abbrev); + Record.clear(); +} + +static void WriteDIDerivedType(const DIDerivedType *N, + const ValueEnumerator &VE, + BitstreamWriter &Stream, + SmallVectorImpl<uint64_t> &Record, + unsigned Abbrev) { + Record.push_back(N->isDistinct()); + Record.push_back(N->getTag()); + Record.push_back(VE.getMetadataOrNullID(N->getRawName())); + Record.push_back(VE.getMetadataOrNullID(N->getFile())); + Record.push_back(N->getLine()); + Record.push_back(VE.getMetadataOrNullID(N->getScope())); + Record.push_back(VE.getMetadataOrNullID(N->getBaseType())); + Record.push_back(N->getSizeInBits()); + Record.push_back(N->getAlignInBits()); + Record.push_back(N->getOffsetInBits()); + Record.push_back(N->getFlags()); + Record.push_back(VE.getMetadataOrNullID(N->getExtraData())); + + Stream.EmitRecord(bitc::METADATA_DERIVED_TYPE, Record, Abbrev); + Record.clear(); +} + +static void WriteDICompositeType(const DICompositeType *N, + const ValueEnumerator &VE, + BitstreamWriter &Stream, + SmallVectorImpl<uint64_t> &Record, + unsigned Abbrev) { + Record.push_back(N->isDistinct()); + Record.push_back(N->getTag()); + Record.push_back(VE.getMetadataOrNullID(N->getRawName())); + Record.push_back(VE.getMetadataOrNullID(N->getFile())); + Record.push_back(N->getLine()); + Record.push_back(VE.getMetadataOrNullID(N->getScope())); + Record.push_back(VE.getMetadataOrNullID(N->getBaseType())); + Record.push_back(N->getSizeInBits()); + Record.push_back(N->getAlignInBits()); + Record.push_back(N->getOffsetInBits()); + Record.push_back(N->getFlags()); + Record.push_back(VE.getMetadataOrNullID(N->getElements().get())); + Record.push_back(N->getRuntimeLang()); + Record.push_back(VE.getMetadataOrNullID(N->getVTableHolder())); + Record.push_back(VE.getMetadataOrNullID(N->getTemplateParams().get())); + Record.push_back(VE.getMetadataOrNullID(N->getRawIdentifier())); + + Stream.EmitRecord(bitc::METADATA_COMPOSITE_TYPE, Record, Abbrev); + Record.clear(); +} + +static void WriteDISubroutineType(const DISubroutineType *N, + const ValueEnumerator &VE, + BitstreamWriter &Stream, + SmallVectorImpl<uint64_t> &Record, + unsigned Abbrev) { + Record.push_back(N->isDistinct()); + Record.push_back(N->getFlags()); + Record.push_back(VE.getMetadataOrNullID(N->getTypeArray().get())); + + Stream.EmitRecord(bitc::METADATA_SUBROUTINE_TYPE, Record, Abbrev); + Record.clear(); +} + +static void WriteDIFile(const DIFile *N, const ValueEnumerator &VE, + BitstreamWriter &Stream, + SmallVectorImpl<uint64_t> &Record, unsigned Abbrev) { + Record.push_back(N->isDistinct()); + Record.push_back(VE.getMetadataOrNullID(N->getRawFilename())); + Record.push_back(VE.getMetadataOrNullID(N->getRawDirectory())); + + Stream.EmitRecord(bitc::METADATA_FILE, Record, Abbrev); + Record.clear(); +} + +static void WriteDICompileUnit(const DICompileUnit *N, + const ValueEnumerator &VE, + BitstreamWriter &Stream, + SmallVectorImpl<uint64_t> &Record, + unsigned Abbrev) { + Record.push_back(N->isDistinct()); + Record.push_back(N->getSourceLanguage()); + Record.push_back(VE.getMetadataOrNullID(N->getFile())); + Record.push_back(VE.getMetadataOrNullID(N->getRawProducer())); + Record.push_back(N->isOptimized()); + Record.push_back(VE.getMetadataOrNullID(N->getRawFlags())); + Record.push_back(N->getRuntimeVersion()); + Record.push_back(VE.getMetadataOrNullID(N->getRawSplitDebugFilename())); + Record.push_back(N->getEmissionKind()); + Record.push_back(VE.getMetadataOrNullID(N->getEnumTypes().get())); + Record.push_back(VE.getMetadataOrNullID(N->getRetainedTypes().get())); + Record.push_back(VE.getMetadataOrNullID(N->getSubprograms().get())); + Record.push_back(VE.getMetadataOrNullID(N->getGlobalVariables().get())); + Record.push_back(VE.getMetadataOrNullID(N->getImportedEntities().get())); + Record.push_back(N->getDWOId()); + + Stream.EmitRecord(bitc::METADATA_COMPILE_UNIT, Record, Abbrev); + Record.clear(); +} + +static void WriteDISubprogram(const DISubprogram *N, const ValueEnumerator &VE, + BitstreamWriter &Stream, + SmallVectorImpl<uint64_t> &Record, + unsigned Abbrev) { + Record.push_back(N->isDistinct()); + Record.push_back(VE.getMetadataOrNullID(N->getScope())); + Record.push_back(VE.getMetadataOrNullID(N->getRawName())); + Record.push_back(VE.getMetadataOrNullID(N->getRawLinkageName())); + Record.push_back(VE.getMetadataOrNullID(N->getFile())); + Record.push_back(N->getLine()); + Record.push_back(VE.getMetadataOrNullID(N->getType())); + Record.push_back(N->isLocalToUnit()); + Record.push_back(N->isDefinition()); + Record.push_back(N->getScopeLine()); + Record.push_back(VE.getMetadataOrNullID(N->getContainingType())); + Record.push_back(N->getVirtuality()); + Record.push_back(N->getVirtualIndex()); + Record.push_back(N->getFlags()); + Record.push_back(N->isOptimized()); + Record.push_back(VE.getMetadataOrNullID(N->getRawFunction())); + Record.push_back(VE.getMetadataOrNullID(N->getTemplateParams().get())); + Record.push_back(VE.getMetadataOrNullID(N->getDeclaration())); + Record.push_back(VE.getMetadataOrNullID(N->getVariables().get())); + + Stream.EmitRecord(bitc::METADATA_SUBPROGRAM, Record, Abbrev); + Record.clear(); +} + +static void WriteDILexicalBlock(const DILexicalBlock *N, + const ValueEnumerator &VE, + BitstreamWriter &Stream, + SmallVectorImpl<uint64_t> &Record, + unsigned Abbrev) { + Record.push_back(N->isDistinct()); + Record.push_back(VE.getMetadataOrNullID(N->getScope())); + Record.push_back(VE.getMetadataOrNullID(N->getFile())); + Record.push_back(N->getLine()); + Record.push_back(N->getColumn()); + + Stream.EmitRecord(bitc::METADATA_LEXICAL_BLOCK, Record, Abbrev); + Record.clear(); +} + +static void WriteDILexicalBlockFile(const DILexicalBlockFile *N, + const ValueEnumerator &VE, + BitstreamWriter &Stream, + SmallVectorImpl<uint64_t> &Record, + unsigned Abbrev) { + Record.push_back(N->isDistinct()); + Record.push_back(VE.getMetadataOrNullID(N->getScope())); + Record.push_back(VE.getMetadataOrNullID(N->getFile())); + Record.push_back(N->getDiscriminator()); + + Stream.EmitRecord(bitc::METADATA_LEXICAL_BLOCK_FILE, Record, Abbrev); + Record.clear(); +} + +static void WriteDINamespace(const DINamespace *N, const ValueEnumerator &VE, + BitstreamWriter &Stream, + SmallVectorImpl<uint64_t> &Record, + unsigned Abbrev) { + Record.push_back(N->isDistinct()); + Record.push_back(VE.getMetadataOrNullID(N->getScope())); + Record.push_back(VE.getMetadataOrNullID(N->getFile())); + Record.push_back(VE.getMetadataOrNullID(N->getRawName())); + Record.push_back(N->getLine()); + + Stream.EmitRecord(bitc::METADATA_NAMESPACE, Record, Abbrev); + Record.clear(); +} + +static void WriteDITemplateTypeParameter(const DITemplateTypeParameter *N, + const ValueEnumerator &VE, + BitstreamWriter &Stream, + SmallVectorImpl<uint64_t> &Record, + unsigned Abbrev) { + Record.push_back(N->isDistinct()); + Record.push_back(VE.getMetadataOrNullID(N->getRawName())); + Record.push_back(VE.getMetadataOrNullID(N->getType())); + + Stream.EmitRecord(bitc::METADATA_TEMPLATE_TYPE, Record, Abbrev); + Record.clear(); +} + +static void WriteDITemplateValueParameter(const DITemplateValueParameter *N, + const ValueEnumerator &VE, + BitstreamWriter &Stream, + SmallVectorImpl<uint64_t> &Record, + unsigned Abbrev) { + Record.push_back(N->isDistinct()); + Record.push_back(N->getTag()); + Record.push_back(VE.getMetadataOrNullID(N->getRawName())); + Record.push_back(VE.getMetadataOrNullID(N->getType())); + Record.push_back(VE.getMetadataOrNullID(N->getValue())); + + Stream.EmitRecord(bitc::METADATA_TEMPLATE_VALUE, Record, Abbrev); + Record.clear(); +} + +static void WriteDIGlobalVariable(const DIGlobalVariable *N, + const ValueEnumerator &VE, + BitstreamWriter &Stream, + SmallVectorImpl<uint64_t> &Record, + unsigned Abbrev) { + Record.push_back(N->isDistinct()); + Record.push_back(VE.getMetadataOrNullID(N->getScope())); + Record.push_back(VE.getMetadataOrNullID(N->getRawName())); + Record.push_back(VE.getMetadataOrNullID(N->getRawLinkageName())); + Record.push_back(VE.getMetadataOrNullID(N->getFile())); + Record.push_back(N->getLine()); + Record.push_back(VE.getMetadataOrNullID(N->getType())); + Record.push_back(N->isLocalToUnit()); + Record.push_back(N->isDefinition()); + Record.push_back(VE.getMetadataOrNullID(N->getRawVariable())); + Record.push_back(VE.getMetadataOrNullID(N->getStaticDataMemberDeclaration())); + + Stream.EmitRecord(bitc::METADATA_GLOBAL_VAR, Record, Abbrev); + Record.clear(); +} + +static void WriteDILocalVariable(const DILocalVariable *N, + const ValueEnumerator &VE, + BitstreamWriter &Stream, + SmallVectorImpl<uint64_t> &Record, + unsigned Abbrev) { + Record.push_back(N->isDistinct()); + Record.push_back(N->getTag()); + Record.push_back(VE.getMetadataOrNullID(N->getScope())); + Record.push_back(VE.getMetadataOrNullID(N->getRawName())); + Record.push_back(VE.getMetadataOrNullID(N->getFile())); + Record.push_back(N->getLine()); + Record.push_back(VE.getMetadataOrNullID(N->getType())); + Record.push_back(N->getArg()); + Record.push_back(N->getFlags()); + + Stream.EmitRecord(bitc::METADATA_LOCAL_VAR, Record, Abbrev); + Record.clear(); +} + +static void WriteDIExpression(const DIExpression *N, const ValueEnumerator &, + BitstreamWriter &Stream, + SmallVectorImpl<uint64_t> &Record, + unsigned Abbrev) { + Record.reserve(N->getElements().size() + 1); + + Record.push_back(N->isDistinct()); + Record.append(N->elements_begin(), N->elements_end()); + + Stream.EmitRecord(bitc::METADATA_EXPRESSION, Record, Abbrev); + Record.clear(); +} + +static void WriteDIObjCProperty(const DIObjCProperty *N, + const ValueEnumerator &VE, + BitstreamWriter &Stream, + SmallVectorImpl<uint64_t> &Record, + unsigned Abbrev) { + Record.push_back(N->isDistinct()); + Record.push_back(VE.getMetadataOrNullID(N->getRawName())); + Record.push_back(VE.getMetadataOrNullID(N->getFile())); + Record.push_back(N->getLine()); + Record.push_back(VE.getMetadataOrNullID(N->getRawSetterName())); + Record.push_back(VE.getMetadataOrNullID(N->getRawGetterName())); + Record.push_back(N->getAttributes()); + Record.push_back(VE.getMetadataOrNullID(N->getType())); + + Stream.EmitRecord(bitc::METADATA_OBJC_PROPERTY, Record, Abbrev); + Record.clear(); +} + +static void WriteDIImportedEntity(const DIImportedEntity *N, + const ValueEnumerator &VE, + BitstreamWriter &Stream, + SmallVectorImpl<uint64_t> &Record, + unsigned Abbrev) { + Record.push_back(N->isDistinct()); + Record.push_back(N->getTag()); + Record.push_back(VE.getMetadataOrNullID(N->getScope())); + Record.push_back(VE.getMetadataOrNullID(N->getEntity())); + Record.push_back(N->getLine()); + Record.push_back(VE.getMetadataOrNullID(N->getRawName())); + + Stream.EmitRecord(bitc::METADATA_IMPORTED_ENTITY, Record, Abbrev); + Record.clear(); +} + static void WriteModuleMetadata(const Module *M, const ValueEnumerator &VE, BitstreamWriter &Stream) { @@ -817,8 +1162,11 @@ static void WriteModuleMetadata(const Module *M, MDSAbbrev = Stream.EmitAbbrev(Abbv); } - unsigned LocAbbrev = 0; - if (VE.hasMDLocation()) { + // Initialize MDNode abbreviations. +#define HANDLE_MDNODE_LEAF(CLASS) unsigned CLASS##Abbrev = 0; +#include "llvm/IR/Metadata.def" + + if (VE.hasDILocation()) { // Abbrev for METADATA_LOCATION. // // Assume the column is usually under 128, and always output the inlined-at @@ -830,7 +1178,23 @@ static void WriteModuleMetadata(const Module *M, Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::VBR, 8)); Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::VBR, 6)); Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::VBR, 6)); - LocAbbrev = Stream.EmitAbbrev(Abbv); + DILocationAbbrev = Stream.EmitAbbrev(Abbv); + } + + if (VE.hasGenericDINode()) { + // Abbrev for METADATA_GENERIC_DEBUG. + // + // Assume the column is usually under 128, and always output the inlined-at + // location (it's never more expensive than building an array size 1). + BitCodeAbbrev *Abbv = new BitCodeAbbrev(); + Abbv->Add(BitCodeAbbrevOp(bitc::METADATA_GENERIC_DEBUG)); + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::Fixed, 1)); + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::VBR, 6)); + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::Fixed, 1)); + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::VBR, 6)); + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::Array)); + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::VBR, 6)); + GenericDINodeAbbrev = Stream.EmitAbbrev(Abbv); } unsigned NameAbbrev = 0; @@ -845,13 +1209,18 @@ static void WriteModuleMetadata(const Module *M, SmallVector<uint64_t, 64> Record; for (const Metadata *MD : MDs) { - if (const MDLocation *Loc = dyn_cast<MDLocation>(MD)) { - WriteMDLocation(Loc, VE, Stream, Record, LocAbbrev); - continue; - } if (const MDNode *N = dyn_cast<MDNode>(MD)) { - WriteMDNode(N, VE, Stream, Record); - continue; + assert(N->isResolved() && "Expected forward references to be resolved"); + + switch (N->getMetadataID()) { + default: + llvm_unreachable("Invalid MDNode subclass"); +#define HANDLE_MDNODE_LEAF(CLASS) \ + case Metadata::CLASS##Kind: \ + Write##CLASS(cast<CLASS>(N), VE, Stream, Record, CLASS##Abbrev); \ + continue; +#include "llvm/IR/Metadata.def" + } } if (const auto *MDC = dyn_cast<ConstantAsMetadata>(MD)) { WriteValueAsMetadata(MDC, VE, Stream, Record); @@ -914,6 +1283,15 @@ static void WriteMetadataAttachment(const Function &F, // Write metadata attachments // METADATA_ATTACHMENT - [m x [value, [n x [id, mdnode]]] SmallVector<std::pair<unsigned, MDNode *>, 4> MDs; + F.getAllMetadata(MDs); + if (!MDs.empty()) { + for (const auto &I : MDs) { + Record.push_back(I.first); + Record.push_back(VE.getMetadataID(I.second)); + } + Stream.EmitRecord(bitc::METADATA_ATTACHMENT, Record, 0); + Record.clear(); + } for (Function::const_iterator BB = F.begin(), E = F.end(); BB != E; ++BB) for (BasicBlock::const_iterator I = BB->begin(), E = BB->end(); @@ -1031,14 +1409,12 @@ static void WriteConstants(unsigned FirstVal, unsigned LastVal, // 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]); + Record.append(AsmStr.begin(), AsmStr.end()); // 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]); + Record.append(ConstraintStr.begin(), ConstraintStr.end()); Stream.EmitRecord(bitc::CST_CODE_INLINEASM, Record); Record.clear(); continue; @@ -1162,15 +1538,18 @@ static void WriteConstants(unsigned FirstVal, unsigned LastVal, Record.push_back(Flags); } break; - case Instruction::GetElementPtr: + case Instruction::GetElementPtr: { Code = bitc::CST_CODE_CE_GEP; - if (cast<GEPOperator>(C)->isInBounds()) + const auto *GO = cast<GEPOperator>(C); + if (GO->isInBounds()) Code = bitc::CST_CODE_CE_INBOUNDS_GEP; + Record.push_back(VE.getTypeID(GO->getSourceElementType())); 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))); } break; + } case Instruction::Select: Code = bitc::CST_CODE_CE_SELECT; Record.push_back(VE.getValueID(C->getOperand(0))); @@ -1316,19 +1695,21 @@ static void WriteInstruction(const Instruction &I, unsigned InstID, } break; - case Instruction::GetElementPtr: + case Instruction::GetElementPtr: { Code = bitc::FUNC_CODE_INST_GEP; - if (cast<GEPOperator>(&I)->isInBounds()) - Code = bitc::FUNC_CODE_INST_INBOUNDS_GEP; + AbbrevToUse = FUNCTION_INST_GEP_ABBREV; + auto &GEPInst = cast<GetElementPtrInst>(I); + Vals.push_back(GEPInst.isInBounds()); + Vals.push_back(VE.getTypeID(GEPInst.getSourceElementType())); 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<ExtractValueInst>(&I); - for (const unsigned *i = EVI->idx_begin(), *e = EVI->idx_end(); i != e; ++i) - Vals.push_back(*i); + Vals.append(EVI->idx_begin(), EVI->idx_end()); break; } case Instruction::InsertValue: { @@ -1336,8 +1717,7 @@ static void WriteInstruction(const Instruction &I, unsigned InstID, PushValueAndType(I.getOperand(0), InstID, Vals, VE); PushValueAndType(I.getOperand(1), InstID, Vals, VE); const InsertValueInst *IVI = cast<InsertValueInst>(&I); - for (const unsigned *i = IVI->idx_begin(), *e = IVI->idx_end(); i != e; ++i) - Vals.push_back(*i); + Vals.append(IVI->idx_begin(), IVI->idx_end()); break; } case Instruction::Select: @@ -1423,15 +1803,15 @@ static void WriteInstruction(const Instruction &I, unsigned InstID, case Instruction::Invoke: { const InvokeInst *II = cast<InvokeInst>(&I); - const Value *Callee(II->getCalledValue()); - PointerType *PTy = cast<PointerType>(Callee->getType()); - FunctionType *FTy = cast<FunctionType>(PTy->getElementType()); + const Value *Callee = II->getCalledValue(); + FunctionType *FTy = II->getFunctionType(); Code = bitc::FUNC_CODE_INST_INVOKE; Vals.push_back(VE.getAttributeID(II->getAttributes())); - Vals.push_back(II->getCallingConv()); + Vals.push_back(II->getCallingConv() | 1 << 13); Vals.push_back(VE.getValueID(II->getNormalDest())); Vals.push_back(VE.getValueID(II->getUnwindDest())); + Vals.push_back(VE.getTypeID(FTy)); PushValueAndType(Callee, InstID, Vals, VE); // Emit value #'s for the fixed parameters. @@ -1492,15 +1872,16 @@ static void WriteInstruction(const Instruction &I, unsigned InstID, case Instruction::Alloca: { Code = bitc::FUNC_CODE_INST_ALLOCA; - Vals.push_back(VE.getTypeID(I.getType())); + const AllocaInst &AI = cast<AllocaInst>(I); + Vals.push_back(VE.getTypeID(AI.getAllocatedType())); Vals.push_back(VE.getTypeID(I.getOperand(0)->getType())); Vals.push_back(VE.getValueID(I.getOperand(0))); // size. - const AllocaInst &AI = cast<AllocaInst>(I); unsigned AlignRecord = Log2_32(AI.getAlignment()) + 1; assert(Log2_32(Value::MaximumAlignment) + 1 < 1 << 5 && "not enough bits for maximum alignment"); assert(AlignRecord < 1 << 5 && "alignment greater than 1 << 64"); AlignRecord |= AI.isUsedWithInAlloca() << 5; + AlignRecord |= 1 << 6; Vals.push_back(AlignRecord); break; } @@ -1514,6 +1895,7 @@ static void WriteInstruction(const Instruction &I, unsigned InstID, if (!PushValueAndType(I.getOperand(0), InstID, Vals, VE)) // ptr AbbrevToUse = FUNCTION_INST_LOAD_ABBREV; } + Vals.push_back(VE.getTypeID(I.getType())); Vals.push_back(Log2_32(cast<LoadInst>(I).getAlignment())+1); Vals.push_back(cast<LoadInst>(I).isVolatile()); if (cast<LoadInst>(I).isAtomic()) { @@ -1527,7 +1909,7 @@ static void WriteInstruction(const Instruction &I, unsigned InstID, else Code = bitc::FUNC_CODE_INST_STORE; PushValueAndType(I.getOperand(1), InstID, Vals, VE); // ptrty + ptr - pushValue(I.getOperand(0), InstID, Vals, VE); // val. + PushValueAndType(I.getOperand(0), InstID, Vals, VE); // valty + val Vals.push_back(Log2_32(cast<StoreInst>(I).getAlignment())+1); Vals.push_back(cast<StoreInst>(I).isVolatile()); if (cast<StoreInst>(I).isAtomic()) { @@ -1538,7 +1920,7 @@ static void WriteInstruction(const Instruction &I, unsigned InstID, case Instruction::AtomicCmpXchg: Code = bitc::FUNC_CODE_INST_CMPXCHG; PushValueAndType(I.getOperand(0), InstID, Vals, VE); // ptrty + ptr - pushValue(I.getOperand(1), InstID, Vals, VE); // cmp. + PushValueAndType(I.getOperand(1), InstID, Vals, VE); // cmp. pushValue(I.getOperand(2), InstID, Vals, VE); // newval. Vals.push_back(cast<AtomicCmpXchgInst>(I).isVolatile()); Vals.push_back(GetEncodedOrdering( @@ -1567,14 +1949,14 @@ static void WriteInstruction(const Instruction &I, unsigned InstID, break; case Instruction::Call: { const CallInst &CI = cast<CallInst>(I); - PointerType *PTy = cast<PointerType>(CI.getCalledValue()->getType()); - FunctionType *FTy = cast<FunctionType>(PTy->getElementType()); + FunctionType *FTy = CI.getFunctionType(); Code = bitc::FUNC_CODE_INST_CALL; Vals.push_back(VE.getAttributeID(CI.getAttributes())); Vals.push_back((CI.getCallingConv() << 1) | unsigned(CI.isTailCall()) | - unsigned(CI.isMustTailCall()) << 14); + unsigned(CI.isMustTailCall()) << 14 | 1 << 15); + Vals.push_back(VE.getTypeID(FTy)); PushValueAndType(CI.getCalledValue(), InstID, Vals, VE); // Callee // Emit value #'s for the fixed parameters. @@ -1673,15 +2055,16 @@ static void WriteUseList(ValueEnumerator &VE, UseListOrder &&Order, else Code = bitc::USELIST_CODE_DEFAULT; - SmallVector<uint64_t, 64> Record; - for (unsigned I : Order.Shuffle) - Record.push_back(I); + SmallVector<uint64_t, 64> Record(Order.Shuffle.begin(), Order.Shuffle.end()); Record.push_back(VE.getValueID(Order.V)); Stream.EmitRecord(Code, Record); } static void WriteUseListBlock(const Function *F, ValueEnumerator &VE, BitstreamWriter &Stream) { + assert(VE.shouldPreserveUseListOrder() && + "Expected to be preserving use-list order"); + auto hasMore = [&]() { return !VE.UseListOrders.empty() && VE.UseListOrders.back().F == F; }; @@ -1722,9 +2105,9 @@ static void WriteFunction(const Function &F, ValueEnumerator &VE, // Keep a running idea of what the instruction ID is. unsigned InstID = CstEnd; - bool NeedsMetadataAttachment = false; + bool NeedsMetadataAttachment = F.hasMetadata(); - DebugLoc LastDL; + DILocation *LastDL = nullptr; // Finally, emit all the instructions, in order. for (Function::const_iterator BB = F.begin(), E = F.end(); BB != E; ++BB) @@ -1739,26 +2122,24 @@ static void WriteFunction(const Function &F, ValueEnumerator &VE, NeedsMetadataAttachment |= I->hasMetadataOtherThanDebugLoc(); // If the instruction has a debug location, emit it. - DebugLoc DL = I->getDebugLoc(); - if (DL.isUnknown()) { - // nothing todo. - } else if (DL == LastDL) { + DILocation *DL = I->getDebugLoc(); + if (!DL) + continue; + + if (DL == LastDL) { // Just repeat the same debug loc as last time. Stream.EmitRecord(bitc::FUNC_CODE_DEBUG_LOC_AGAIN, Vals); - } else { - MDNode *Scope, *IA; - DL.getScopeAndInlinedAt(Scope, IA, I->getContext()); - assert(Scope && "Expected valid scope"); - - Vals.push_back(DL.getLine()); - Vals.push_back(DL.getCol()); - Vals.push_back(Scope ? VE.getMetadataID(Scope) + 1 : 0); - Vals.push_back(IA ? VE.getMetadataID(IA) + 1 : 0); - Stream.EmitRecord(bitc::FUNC_CODE_DEBUG_LOC, Vals); - Vals.clear(); - - LastDL = DL; + continue; } + + Vals.push_back(DL->getLine()); + Vals.push_back(DL->getColumn()); + Vals.push_back(VE.getMetadataOrNullID(DL->getScope())); + Vals.push_back(VE.getMetadataOrNullID(DL->getInlinedAt())); + Stream.EmitRecord(bitc::FUNC_CODE_DEBUG_LOC, Vals); + Vals.clear(); + + LastDL = DL; } // Emit names for all the instructions etc. @@ -1766,7 +2147,7 @@ static void WriteFunction(const Function &F, ValueEnumerator &VE, if (NeedsMetadataAttachment) WriteMetadataAttachment(F, VE, Stream); - if (shouldPreserveBitcodeUseListOrder()) + if (VE.shouldPreserveUseListOrder()) WriteUseListBlock(&F, VE, Stream); VE.purgeFunction(); Stream.ExitBlock(); @@ -1827,7 +2208,7 @@ static void WriteBlockInfo(const ValueEnumerator &VE, BitstreamWriter &Stream) { BitCodeAbbrev *Abbv = new BitCodeAbbrev(); Abbv->Add(BitCodeAbbrevOp(bitc::CST_CODE_SETTYPE)); Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::Fixed, - Log2_32_Ceil(VE.getTypes().size()+1))); + VE.computeBitsRequiredForTypeIndicies())); if (Stream.EmitBlockInfoAbbrev(bitc::CONSTANTS_BLOCK_ID, Abbv) != CONSTANTS_SETTYPE_ABBREV) llvm_unreachable("Unexpected abbrev ordering!"); @@ -1847,7 +2228,7 @@ static void WriteBlockInfo(const ValueEnumerator &VE, BitstreamWriter &Stream) { 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))); + VE.computeBitsRequiredForTypeIndicies())); Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::VBR, 8)); // value id if (Stream.EmitBlockInfoAbbrev(bitc::CONSTANTS_BLOCK_ID, @@ -1868,6 +2249,8 @@ static void WriteBlockInfo(const ValueEnumerator &VE, BitstreamWriter &Stream) { BitCodeAbbrev *Abbv = new BitCodeAbbrev(); Abbv->Add(BitCodeAbbrevOp(bitc::FUNC_CODE_INST_LOAD)); Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::VBR, 6)); // Ptr + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::Fixed, // dest ty + VE.computeBitsRequiredForTypeIndicies())); Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::VBR, 4)); // Align Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::Fixed, 1)); // volatile if (Stream.EmitBlockInfoAbbrev(bitc::FUNCTION_BLOCK_ID, @@ -1900,7 +2283,7 @@ static void WriteBlockInfo(const ValueEnumerator &VE, BitstreamWriter &Stream) { 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))); + VE.computeBitsRequiredForTypeIndicies())); Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::Fixed, 4)); // opc if (Stream.EmitBlockInfoAbbrev(bitc::FUNCTION_BLOCK_ID, Abbv) != FUNCTION_INST_CAST_ABBREV) @@ -1929,12 +2312,25 @@ static void WriteBlockInfo(const ValueEnumerator &VE, BitstreamWriter &Stream) { Abbv) != FUNCTION_INST_UNREACHABLE_ABBREV) llvm_unreachable("Unexpected abbrev ordering!"); } + { + BitCodeAbbrev *Abbv = new BitCodeAbbrev(); + Abbv->Add(BitCodeAbbrevOp(bitc::FUNC_CODE_INST_GEP)); + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::Fixed, 1)); + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::Fixed, // dest ty + Log2_32_Ceil(VE.getTypes().size() + 1))); + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::Array)); + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::VBR, 6)); + if (Stream.EmitBlockInfoAbbrev(bitc::FUNCTION_BLOCK_ID, Abbv) != + FUNCTION_INST_GEP_ABBREV) + llvm_unreachable("Unexpected abbrev ordering!"); + } Stream.ExitBlock(); } /// WriteModule - Emit the specified module to the bitstream. -static void WriteModule(const Module *M, BitstreamWriter &Stream) { +static void WriteModule(const Module *M, BitstreamWriter &Stream, + bool ShouldPreserveUseListOrder) { Stream.EnterSubblock(bitc::MODULE_BLOCK_ID, 3); SmallVector<unsigned, 1> Vals; @@ -1943,7 +2339,7 @@ static void WriteModule(const Module *M, BitstreamWriter &Stream) { Stream.EmitRecord(bitc::MODULE_CODE_VERSION, Vals); // Analyze the module, enumerating globals, functions, etc. - ValueEnumerator VE(*M); + ValueEnumerator VE(*M, ShouldPreserveUseListOrder); // Emit blockinfo, which defines the standard abbreviations etc. WriteBlockInfo(VE, Stream); @@ -1976,7 +2372,7 @@ static void WriteModule(const Module *M, BitstreamWriter &Stream) { WriteValueSymbolTable(M->getValueSymbolTable(), VE, Stream); // Emit module-level use-lists. - if (shouldPreserveBitcodeUseListOrder()) + if (VE.shouldPreserveUseListOrder()) WriteUseListBlock(nullptr, VE, Stream); // Emit function bodies. @@ -2062,7 +2458,8 @@ static void EmitDarwinBCHeaderAndTrailer(SmallVectorImpl<char> &Buffer, /// WriteBitcodeToFile - Write the specified module to the specified output /// stream. -void llvm::WriteBitcodeToFile(const Module *M, raw_ostream &Out) { +void llvm::WriteBitcodeToFile(const Module *M, raw_ostream &Out, + bool ShouldPreserveUseListOrder) { SmallVector<char, 0> Buffer; Buffer.reserve(256*1024); @@ -2085,7 +2482,7 @@ void llvm::WriteBitcodeToFile(const Module *M, raw_ostream &Out) { Stream.Emit(0xD, 4); // Emit the module. - WriteModule(M, Stream); + WriteModule(M, Stream, ShouldPreserveUseListOrder); } if (TT.isOSDarwin()) diff --git a/lib/Bitcode/Writer/BitcodeWriterPass.cpp b/lib/Bitcode/Writer/BitcodeWriterPass.cpp index 25456a456217..3165743576ec 100644 --- a/lib/Bitcode/Writer/BitcodeWriterPass.cpp +++ b/lib/Bitcode/Writer/BitcodeWriterPass.cpp @@ -19,22 +19,25 @@ using namespace llvm; PreservedAnalyses BitcodeWriterPass::run(Module &M) { - WriteBitcodeToFile(&M, OS); + WriteBitcodeToFile(&M, OS, ShouldPreserveUseListOrder); return PreservedAnalyses::all(); } namespace { class WriteBitcodePass : public ModulePass { raw_ostream &OS; // raw_ostream to print on + bool ShouldPreserveUseListOrder; + public: static char ID; // Pass identification, replacement for typeid - explicit WriteBitcodePass(raw_ostream &o) - : ModulePass(ID), OS(o) {} + explicit WriteBitcodePass(raw_ostream &o, bool ShouldPreserveUseListOrder) + : ModulePass(ID), OS(o), + ShouldPreserveUseListOrder(ShouldPreserveUseListOrder) {} const char *getPassName() const override { return "Bitcode Writer"; } bool runOnModule(Module &M) override { - WriteBitcodeToFile(&M, OS); + WriteBitcodeToFile(&M, OS, ShouldPreserveUseListOrder); return false; } }; @@ -42,6 +45,7 @@ namespace { char WriteBitcodePass::ID = 0; -ModulePass *llvm::createBitcodeWriterPass(raw_ostream &Str) { - return new WriteBitcodePass(Str); +ModulePass *llvm::createBitcodeWriterPass(raw_ostream &Str, + bool ShouldPreserveUseListOrder) { + return new WriteBitcodePass(Str, ShouldPreserveUseListOrder); } diff --git a/lib/Bitcode/Writer/ValueEnumerator.cpp b/lib/Bitcode/Writer/ValueEnumerator.cpp index 27a63d8fefc6..6c517f5ed8d0 100644 --- a/lib/Bitcode/Writer/ValueEnumerator.cpp +++ b/lib/Bitcode/Writer/ValueEnumerator.cpp @@ -15,6 +15,7 @@ #include "llvm/ADT/STLExtras.h" #include "llvm/ADT/SmallPtrSet.h" #include "llvm/IR/Constants.h" +#include "llvm/IR/DebugInfoMetadata.h" #include "llvm/IR/DerivedTypes.h" #include "llvm/IR/Instructions.h" #include "llvm/IR/Module.h" @@ -282,9 +283,11 @@ static bool isIntOrIntVectorValue(const std::pair<const Value*, unsigned> &V) { return V.first->getType()->isIntOrIntVectorTy(); } -ValueEnumerator::ValueEnumerator(const Module &M) - : HasMDString(false), HasMDLocation(false) { - if (shouldPreserveBitcodeUseListOrder()) +ValueEnumerator::ValueEnumerator(const Module &M, + bool ShouldPreserveUseListOrder) + : HasMDString(false), HasDILocation(false), HasGenericDINode(false), + ShouldPreserveUseListOrder(ShouldPreserveUseListOrder) { + if (ShouldPreserveUseListOrder) UseListOrders = predictUseListOrder(M); // Enumerate the global variables. @@ -345,6 +348,11 @@ ValueEnumerator::ValueEnumerator(const Module &M) for (const Argument &A : F.args()) EnumerateType(A.getType()); + // Enumerate metadata attached to this function. + F.getAllMetadata(MDs); + for (const auto &I : MDs) + EnumerateMetadata(I.second); + for (const BasicBlock &BB : F) for (const Instruction &I : BB) { for (const Use &Op : I.operands()) { @@ -372,12 +380,10 @@ ValueEnumerator::ValueEnumerator(const Module &M) for (unsigned i = 0, e = MDs.size(); i != e; ++i) EnumerateMetadata(MDs[i].second); - if (!I.getDebugLoc().isUnknown()) { - MDNode *Scope, *IA; - I.getDebugLoc().getScopeAndInlinedAt(Scope, IA, I.getContext()); - if (Scope) EnumerateMetadata(Scope); - if (IA) EnumerateMetadata(IA); - } + // Don't enumerate the location directly -- it has a special record + // type -- but enumerate its operands. + if (DILocation *L = I.getDebugLoc()) + EnumerateMDNodeOperands(L); } } @@ -410,12 +416,6 @@ unsigned ValueEnumerator::getValueID(const Value *V) const { return I->second-1; } -unsigned ValueEnumerator::getMetadataID(const Metadata *MD) const { - auto I = MDValueMap.find(MD); - assert(I != MDValueMap.end() && "Metadata not in slotcalculator!"); - return I->second - 1; -} - void ValueEnumerator::dump() const { print(dbgs(), ValueMap, "Default"); dbgs() << '\n'; @@ -468,7 +468,7 @@ void ValueEnumerator::print(raw_ostream &OS, const MetadataMapType &Map, void ValueEnumerator::OptimizeConstants(unsigned CstStart, unsigned CstEnd) { if (CstStart == CstEnd || CstStart+1 == CstEnd) return; - if (shouldPreserveBitcodeUseListOrder()) + if (ShouldPreserveUseListOrder) // Optimizing constants makes the use-list order difficult to predict. // Disable it for now when trying to preserve the order. return; @@ -548,7 +548,8 @@ void ValueEnumerator::EnumerateMetadata(const Metadata *MD) { EnumerateValue(C->getValue()); HasMDString |= isa<MDString>(MD); - HasMDLocation |= isa<MDLocation>(MD); + HasDILocation |= isa<DILocation>(MD); + HasGenericDINode |= isa<GenericDINode>(MD); // Replace the dummy ID inserted above with the correct one. MDValueMap may // have changed by inserting operands, so we need a fresh lookup here. @@ -807,3 +808,7 @@ unsigned ValueEnumerator::getGlobalBasicBlockID(const BasicBlock *BB) const { IncorporateFunctionInfoGlobalBBIDs(BB->getParent(), GlobalBasicBlockIDs); return getGlobalBasicBlockID(BB); } + +uint64_t ValueEnumerator::computeBitsRequiredForTypeIndicies() const { + return Log2_32_Ceil(getTypes().size() + 1); +} diff --git a/lib/Bitcode/Writer/ValueEnumerator.h b/lib/Bitcode/Writer/ValueEnumerator.h index d363c1be0dda..92d166e3ba92 100644 --- a/lib/Bitcode/Writer/ValueEnumerator.h +++ b/lib/Bitcode/Writer/ValueEnumerator.h @@ -65,7 +65,9 @@ private: typedef DenseMap<const Metadata *, unsigned> MetadataMapType; MetadataMapType MDValueMap; bool HasMDString; - bool HasMDLocation; + bool HasDILocation; + bool HasGenericDINode; + bool ShouldPreserveUseListOrder; typedef DenseMap<AttributeSet, unsigned> AttributeGroupMapType; AttributeGroupMapType AttributeGroupMap; @@ -98,10 +100,10 @@ private: unsigned FirstFuncConstantID; unsigned FirstInstID; - ValueEnumerator(const ValueEnumerator &) LLVM_DELETED_FUNCTION; - void operator=(const ValueEnumerator &) LLVM_DELETED_FUNCTION; + ValueEnumerator(const ValueEnumerator &) = delete; + void operator=(const ValueEnumerator &) = delete; public: - ValueEnumerator(const Module &M); + ValueEnumerator(const Module &M, bool ShouldPreserveUseListOrder); void dump() const; void print(raw_ostream &OS, const ValueMapType &Map, const char *Name) const; @@ -109,10 +111,20 @@ public: const char *Name) const; unsigned getValueID(const Value *V) const; - unsigned getMetadataID(const Metadata *V) const; + unsigned getMetadataID(const Metadata *MD) const { + auto ID = getMetadataOrNullID(MD); + assert(ID != 0 && "Metadata not in slotcalculator!"); + return ID - 1; + } + unsigned getMetadataOrNullID(const Metadata *MD) const { + return MDValueMap.lookup(MD); + } bool hasMDString() const { return HasMDString; } - bool hasMDLocation() const { return HasMDLocation; } + bool hasDILocation() const { return HasDILocation; } + bool hasGenericDINode() const { return HasGenericDINode; } + + bool shouldPreserveUseListOrder() const { return ShouldPreserveUseListOrder; } unsigned getTypeID(Type *T) const { TypeMapType::const_iterator I = TypeMap.find(T); @@ -173,6 +185,7 @@ public: /// void incorporateFunction(const Function &F); void purgeFunction(); + uint64_t computeBitsRequiredForTypeIndicies() const; private: void OptimizeConstants(unsigned CstStart, unsigned CstEnd); |