From 1c9687eed2f17fa478751d6737d9ef1bd4075b22 Mon Sep 17 00:00:00 2001 From: Rafael Espindola Date: Fri, 4 Jul 2014 18:40:36 +0000 Subject: [PATCH] Implement LTOModule on top of IRObjectFile. IRObjectFile provides all the logic for producing mangled names and getting symbols from inline assembly. LTOModule then adds logic for linking specific tasks, like constructing llvm.compiler_user or extracting linker options from the bitcode. The rule of the thumb is that IRObjectFile has the functionality that is needed by both LTO and llvm-ar. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@212349 91177308-0d34-0410-b5e6-96231b3b80d8 --- include/llvm/LTO/LTOModule.h | 46 +++---- include/llvm/Object/IRObjectFile.h | 10 +- lib/LTO/LTOCodeGenerator.cpp | 2 +- lib/LTO/LTOModule.cpp | 201 +++++++++++------------------ lib/Object/IRObjectFile.cpp | 26 ++-- 5 files changed, 117 insertions(+), 168 deletions(-) diff --git a/include/llvm/LTO/LTOModule.h b/include/llvm/LTO/LTOModule.h index 8208b2fb260..c43846a6405 100644 --- a/include/llvm/LTO/LTOModule.h +++ b/include/llvm/LTO/LTOModule.h @@ -16,10 +16,10 @@ #include "llvm-c/lto.h" #include "llvm/ADT/StringMap.h" -#include "llvm/IR/Mangler.h" #include "llvm/IR/Module.h" #include "llvm/MC/MCContext.h" #include "llvm/MC/MCObjectFileInfo.h" +#include "llvm/Object/IRObjectFile.h" #include "llvm/Target/TargetMachine.h" #include #include @@ -46,9 +46,8 @@ private: const GlobalValue *symbol; }; - std::unique_ptr _module; + std::unique_ptr IRFile; std::unique_ptr _target; - MCObjectFileInfo ObjFileInfo; StringSet _linkeropt_strings; std::vector _deplibs; std::vector _linkeropts; @@ -58,12 +57,8 @@ private: StringSet _defines; StringMap _undefines; std::vector _asm_undefines; - MCContext _context; - // Use mangler to add GlobalPrefix to names to match linker names. - Mangler _mangler; - - LTOModule(std::unique_ptr M, TargetMachine *TM); + LTOModule(std::unique_ptr Obj, TargetMachine *TM); public: /// Returns 'true' if the file or memory contents is LLVM bitcode. @@ -100,14 +95,21 @@ public: TargetOptions options, std::string &errMsg, StringRef path = ""); + const Module &getModule() const { + return const_cast(this)->getModule(); + } + Module &getModule() { + return IRFile->getModule(); + } + /// Return the Module's target triple. const std::string &getTargetTriple() { - return _module->getTargetTriple(); + return getModule().getTargetTriple(); } /// Set the Module's target triple. void setTargetTriple(StringRef Triple) { - _module->setTargetTriple(Triple); + getModule().setTargetTriple(Triple); } /// Get the number of symbols @@ -153,9 +155,6 @@ public: return nullptr; } - /// Return the Module. - Module *getLLVVMModule() { return _module.get(); } - const std::vector &getAsmUndefinedRefs() { return _asm_undefines; } @@ -170,23 +169,20 @@ private: bool parseSymbols(std::string &errMsg); /// Add a symbol which isn't defined just yet to a list to be resolved later. - void addPotentialUndefinedSymbol(const GlobalValue *dcl, bool isFunc); + void addPotentialUndefinedSymbol(const object::BasicSymbolRef &Sym, + bool isFunc); /// Add a defined symbol to the list. - void addDefinedSymbol(const char *Name, const GlobalValue *Def, - bool IsFunction); - - /// Add a function symbol as defined to the list. - void addDefinedFunctionSymbol(const Function *f); - void addDefinedFunctionSymbol(const char *Name, const Function *F); + void addDefinedSymbol(const char *Name, const GlobalValue *def, + bool isFunction); /// Add a data symbol as defined to the list. - void addDefinedDataSymbol(const GlobalValue *v); - void addDefinedDataSymbol(const char *Name, const GlobalValue *V); + void addDefinedDataSymbol(const object::BasicSymbolRef &Sym); + void addDefinedDataSymbol(const char*Name, const GlobalValue *v); - /// Add global symbols from module-level ASM to the defined or undefined - /// lists. - bool addAsmGlobalSymbols(std::string &errMsg); + /// Add a function symbol as defined to the list. + void addDefinedFunctionSymbol(const object::BasicSymbolRef &Sym); + void addDefinedFunctionSymbol(const char *Name, const Function *F); /// Add a global symbol from module-level ASM to the defined list. void addAsmGlobalSymbol(const char *, lto_symbol_attributes scope); diff --git a/include/llvm/Object/IRObjectFile.h b/include/llvm/Object/IRObjectFile.h index bc69de40cb5..b33cc263b9b 100644 --- a/include/llvm/Object/IRObjectFile.h +++ b/include/llvm/Object/IRObjectFile.h @@ -28,8 +28,7 @@ class IRObjectFile : public SymbolicFile { std::vector> AsmSymbols; public: - IRObjectFile(std::unique_ptr Object, std::error_code &EC, - LLVMContext &Context); + IRObjectFile(std::unique_ptr Object, std::unique_ptr M); ~IRObjectFile(); void moveSymbolNext(DataRefImpl &Symb) const override; std::error_code printSymbolName(raw_ostream &OS, @@ -39,6 +38,13 @@ public: basic_symbol_iterator symbol_begin_impl() const override; basic_symbol_iterator symbol_end_impl() const override; + const Module &getModule() const { + return const_cast(this)->getModule(); + } + Module &getModule() { + return *M; + } + static inline bool classof(const Binary *v) { return v->isIR(); } diff --git a/lib/LTO/LTOCodeGenerator.cpp b/lib/LTO/LTOCodeGenerator.cpp index a1709f60fff..4b51166a755 100644 --- a/lib/LTO/LTOCodeGenerator.cpp +++ b/lib/LTO/LTOCodeGenerator.cpp @@ -114,7 +114,7 @@ void LTOCodeGenerator::initializeLTOPasses() { } bool LTOCodeGenerator::addModule(LTOModule* mod, std::string& errMsg) { - bool ret = IRLinker.linkInModule(mod->getLLVVMModule(), &errMsg); + bool ret = IRLinker.linkInModule(&mod->getModule(), &errMsg); const std::vector &undefs = mod->getAsmUndefinedRefs(); for (int i = 0, e = undefs.size(); i != e; ++i) diff --git a/lib/LTO/LTOModule.cpp b/lib/LTO/LTOModule.cpp index 45e3a21df72..314831ff6f8 100644 --- a/lib/LTO/LTOModule.cpp +++ b/lib/LTO/LTOModule.cpp @@ -28,7 +28,6 @@ #include "llvm/MC/MCSymbol.h" #include "llvm/MC/MCTargetAsmParser.h" #include "llvm/MC/SubtargetFeature.h" -#include "llvm/Object/RecordStreamer.h" #include "llvm/Support/CommandLine.h" #include "llvm/Support/FileSystem.h" #include "llvm/Support/Host.h" @@ -44,15 +43,9 @@ #include using namespace llvm; -LTOModule::LTOModule(std::unique_ptr M, TargetMachine *TM) - : _module(std::move(M)), _target(TM), - _context(_target->getMCAsmInfo(), _target->getRegisterInfo(), - &ObjFileInfo), - _mangler(TM->getDataLayout()) { - ObjFileInfo.InitMCObjectFileInfo(TM->getTargetTriple(), - TM->getRelocationModel(), TM->getCodeModel(), - _context); -} +LTOModule::LTOModule(std::unique_ptr Obj, + llvm::TargetMachine *TM) + : IRFile(std::move(Obj)), _target(TM) {} /// isBitcodeFile - Returns 'true' if the file (or memory contents) is LLVM /// bitcode. @@ -115,17 +108,15 @@ LTOModule *LTOModule::createFromBuffer(const void *mem, size_t length, LTOModule *LTOModule::makeLTOModule(std::unique_ptr Buffer, TargetOptions options, std::string &errMsg) { - // parse bitcode buffer - ErrorOr ModuleOrErr = + ErrorOr MOrErr = getLazyBitcodeModule(Buffer.get(), getGlobalContext()); - if (std::error_code EC = ModuleOrErr.getError()) { + if (std::error_code EC = MOrErr.getError()) { errMsg = EC.message(); return nullptr; } - Buffer.release(); - std::unique_ptr m(ModuleOrErr.get()); + std::unique_ptr M(MOrErr.get()); - std::string TripleStr = m->getTargetTriple(); + std::string TripleStr = M->getTargetTriple(); if (TripleStr.empty()) TripleStr = sys::getDefaultTargetTriple(); llvm::Triple Triple(TripleStr); @@ -153,18 +144,13 @@ LTOModule *LTOModule::makeLTOModule(std::unique_ptr Buffer, TargetMachine *target = march->createTargetMachine(TripleStr, CPU, FeatureStr, options); - m->materializeAllPermanently(); + M->materializeAllPermanently(true); + M->setDataLayout(target->getDataLayout()); - LTOModule *Ret = new LTOModule(std::move(m), target); + std::unique_ptr IRObj( + new object::IRObjectFile(std::move(Buffer), std::move(M))); - // We need a MCContext set up in order to get mangled names of private - // symbols. It is a bit odd that we need to report uses and definitions - // of private symbols, but it does look like ld64 expects to be informed - // of at least the ones with an 'l' prefix. - MCContext &Context = Ret->_context; - const TargetLoweringObjectFile &TLOF = - target->getTargetLowering()->getObjFileLowering(); - const_cast(TLOF).Initialize(Context, *target); + LTOModule *Ret = new LTOModule(std::move(IRObj), target); if (Ret->parseSymbols(errMsg)) { delete Ret; @@ -283,9 +269,14 @@ void LTOModule::addObjCClassRef(const GlobalVariable *clgv) { entry.setValue(info); } -void LTOModule::addDefinedDataSymbol(const GlobalValue *V) { +void LTOModule::addDefinedDataSymbol(const object::BasicSymbolRef &Sym) { SmallString<64> Buffer; - _target->getNameWithPrefix(Buffer, V, _mangler); + { + raw_svector_ostream OS(Buffer); + Sym.printName(OS); + } + + const GlobalValue *V = IRFile->getSymbolGV(Sym.getRawDataRefImpl()); addDefinedDataSymbol(Buffer.c_str(), V); } @@ -339,9 +330,15 @@ void LTOModule::addDefinedDataSymbol(const char *Name, const GlobalValue *v) { } } -void LTOModule::addDefinedFunctionSymbol(const Function *F) { +void LTOModule::addDefinedFunctionSymbol(const object::BasicSymbolRef &Sym) { SmallString<64> Buffer; - _target->getNameWithPrefix(Buffer, F, _mangler); + { + raw_svector_ostream OS(Buffer); + Sym.printName(OS); + } + + const Function *F = + cast(IRFile->getSymbolGV(Sym.getRawDataRefImpl())); addDefinedFunctionSymbol(Buffer.c_str(), F); } @@ -376,10 +373,6 @@ static bool canBeHidden(const GlobalValue *GV) { void LTOModule::addDefinedSymbol(const char *Name, const GlobalValue *def, bool isFunction) { - // ignore all llvm.* symbols - if (def->getName().startswith("llvm.")) - return; - // set alignment part log2() can have rounding errors uint32_t align = def->getAlignment(); uint32_t attr = align ? countTrailingZeros(align) : 0; @@ -499,20 +492,14 @@ void LTOModule::addAsmGlobalSymbolUndef(const char *name) { entry.setValue(info); } -/// addPotentialUndefinedSymbol - Add a symbol which isn't defined just yet to a -/// list to be resolved later. -void -LTOModule::addPotentialUndefinedSymbol(const GlobalValue *decl, bool isFunc) { - // ignore all llvm.* symbols - if (decl->getName().startswith("llvm.")) - return; - - // ignore all aliases - if (isa(decl)) - return; - +/// Add a symbol which isn't defined just yet to a list to be resolved later. +void LTOModule::addPotentialUndefinedSymbol(const object::BasicSymbolRef &Sym, + bool isFunc) { SmallString<64> name; - _target->getNameWithPrefix(name, decl, _mangler); + { + raw_svector_ostream OS(name); + Sym.printName(OS); + } StringMap::value_type &entry = _undefines.GetOrCreateValue(name); @@ -525,6 +512,8 @@ LTOModule::addPotentialUndefinedSymbol(const GlobalValue *decl, bool isFunc) { info.name = entry.getKey().data(); + const GlobalValue *decl = IRFile->getSymbolGV(Sym.getRawDataRefImpl()); + if (decl->hasExternalWeakLinkage()) info.attributes = LTO_SYMBOL_DEFINITION_WEAKUNDEF; else @@ -536,90 +525,50 @@ LTOModule::addPotentialUndefinedSymbol(const GlobalValue *decl, bool isFunc) { entry.setValue(info); } -/// addAsmGlobalSymbols - Add global symbols from module-level ASM to the -/// defined or undefined lists. -bool LTOModule::addAsmGlobalSymbols(std::string &errMsg) { - const std::string &inlineAsm = _module->getModuleInlineAsm(); - if (inlineAsm.empty()) - return false; - - std::unique_ptr Streamer(new RecordStreamer(_context)); - MemoryBuffer *Buffer = MemoryBuffer::getMemBuffer(inlineAsm); - SourceMgr SrcMgr; - SrcMgr.AddNewSourceBuffer(Buffer, SMLoc()); - std::unique_ptr Parser( - createMCAsmParser(SrcMgr, _context, *Streamer, *_target->getMCAsmInfo())); - const Target &T = _target->getTarget(); - std::unique_ptr MCII(T.createMCInstrInfo()); - std::unique_ptr STI(T.createMCSubtargetInfo( - _target->getTargetTriple(), _target->getTargetCPU(), - _target->getTargetFeatureString())); - std::unique_ptr TAP( - T.createMCAsmParser(*STI, *Parser.get(), *MCII, - _target->Options.MCOptions)); - if (!TAP) { - errMsg = "target " + std::string(T.getName()) + - " does not define AsmParser."; - return true; - } - - Parser->setTargetParser(*TAP); - if (Parser->Run(false)) - return true; - - for (auto &KV : *Streamer) { - StringRef Key = KV.first(); - RecordStreamer::State Value = KV.second; - if (Value == RecordStreamer::DefinedGlobal) - addAsmGlobalSymbol(Key.data(), LTO_SYMBOL_SCOPE_DEFAULT); - else if (Value == RecordStreamer::Defined) - addAsmGlobalSymbol(Key.data(), LTO_SYMBOL_SCOPE_INTERNAL); - else if (Value == RecordStreamer::Global || - Value == RecordStreamer::Used) - addAsmGlobalSymbolUndef(Key.data()); - } - - return false; -} - -/// isDeclaration - Return 'true' if the global value is a declaration. -static bool isDeclaration(const GlobalValue &V) { - if (V.hasAvailableExternallyLinkage()) - return true; - - if (V.isMaterializable()) - return false; - - return V.isDeclaration(); -} - /// parseSymbols - Parse the symbols from the module and model-level ASM and add /// them to either the defined or undefined lists. bool LTOModule::parseSymbols(std::string &errMsg) { - // add functions - for (Module::iterator f = _module->begin(), e = _module->end(); f != e; ++f) { - if (isDeclaration(*f)) - addPotentialUndefinedSymbol(f, true); - else - addDefinedFunctionSymbol(f); - } + for (auto &Sym : IRFile->symbols()) { + const GlobalValue *GV = IRFile->getSymbolGV(Sym.getRawDataRefImpl()); + uint32_t Flags = Sym.getFlags(); + bool IsUndefined = Flags & object::BasicSymbolRef::SF_Undefined; + + if (!GV) { + SmallString<64> Buffer; + { + raw_svector_ostream OS(Buffer); + Sym.printName(OS); + } + const char *Name = Buffer.c_str(); + + if (IsUndefined) + addAsmGlobalSymbolUndef(Name); + else if (Flags & object::BasicSymbolRef::SF_Global) + addAsmGlobalSymbol(Name, LTO_SYMBOL_SCOPE_DEFAULT); + else + addAsmGlobalSymbol(Name, LTO_SYMBOL_SCOPE_INTERNAL); + continue; + } - // add data - for (Module::global_iterator v = _module->global_begin(), - e = _module->global_end(); v != e; ++v) { - if (isDeclaration(*v)) - addPotentialUndefinedSymbol(v, false); - else - addDefinedDataSymbol(v); - } + auto *F = dyn_cast(GV); + if (IsUndefined) { + addPotentialUndefinedSymbol(Sym, F != nullptr); + continue; + } - // add asm globals - if (addAsmGlobalSymbols(errMsg)) - return true; + if (F) { + addDefinedFunctionSymbol(Sym); + continue; + } - // add aliases - for (const auto &Alias : _module->aliases()) - addDefinedDataSymbol(&Alias); + if (isa(GV)) { + addDefinedDataSymbol(Sym); + continue; + } + + assert(isa(GV)); + addDefinedDataSymbol(Sym); + } // make symbols for all undefines for (StringMap::iterator u =_undefines.begin(), @@ -637,7 +586,7 @@ bool LTOModule::parseSymbols(std::string &errMsg) { /// parseMetadata - Parse metadata from the module void LTOModule::parseMetadata() { // Linker Options - if (Value *Val = _module->getModuleFlag("Linker Options")) { + if (Value *Val = getModule().getModuleFlag("Linker Options")) { MDNode *LinkerOptions = cast(Val); for (unsigned i = 0, e = LinkerOptions->getNumOperands(); i != e; ++i) { MDNode *MDOptions = cast(LinkerOptions->getOperand(i)); diff --git a/lib/Object/IRObjectFile.cpp b/lib/Object/IRObjectFile.cpp index 0639c71d1b0..e927447739e 100644 --- a/lib/Object/IRObjectFile.cpp +++ b/lib/Object/IRObjectFile.cpp @@ -33,14 +33,8 @@ using namespace llvm; using namespace object; IRObjectFile::IRObjectFile(std::unique_ptr Object, - std::error_code &EC, LLVMContext &Context) - : SymbolicFile(Binary::ID_IR, std::move(Object)) { - ErrorOr MOrErr = getLazyBitcodeModule(Data.get(), Context); - if ((EC = MOrErr.getError())) - return; - - M.reset(MOrErr.get()); - + std::unique_ptr Mod) + : SymbolicFile(Binary::ID_IR, std::move(Object)), M(std::move(Mod)) { // If we have a DataLayout, setup a mangler. const DataLayout *DL = M->getDataLayout(); if (!DL) @@ -119,7 +113,11 @@ IRObjectFile::IRObjectFile(std::unique_ptr Object, } } -IRObjectFile::~IRObjectFile() { M->getMaterializer()->releaseBuffer(); } +IRObjectFile::~IRObjectFile() { + GVMaterializer *GVM = M->getMaterializer(); + if (GVM) + GVM->releaseBuffer(); + } static const GlobalValue *getGV(DataRefImpl &Symb) { if ((Symb.p & 3) == 3) @@ -275,10 +273,10 @@ basic_symbol_iterator IRObjectFile::symbol_end_impl() const { ErrorOr llvm::object::IRObjectFile::createIRObjectFile( std::unique_ptr Object, LLVMContext &Context) { - std::error_code EC; - std::unique_ptr Ret( - new IRObjectFile(std::move(Object), EC, Context)); - if (EC) + ErrorOr MOrErr = getLazyBitcodeModule(Object.get(), Context); + if (std::error_code EC = MOrErr.getError()) return EC; - return Ret.release(); + + std::unique_ptr M(MOrErr.get()); + return new IRObjectFile(std::move(Object), std::move(M)); } -- 2.34.1